https://aiml.sh/New lessons in aiml.sh courses: runnable code, tests and failure cases for AI applications.enhttps://aiml.sh/courses/building-ai-systems/admit-decoding-requests-between-token-steps/aiml.sh:admit-decoding-requests-between-token-stepsCompute last-position logits for unequal-length prefixes in one padded tensor batch. Execute a focused local CPU experiment and inspect its raw result.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/llm-inference/autoscale-gpus-with-slow-cold-starts-and-roll-out-model-updates-safely/aiml.sh:autoscale-gpus-with-slow-cold-starts-and-roll-out-model-updates-safelyModel replica startup and finite standby capacity, then evaluate canary gates. Distinguish requested capacity from ready capacity and compare tradeoffs.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/llm-inference/benchmark-streaming-inference/aiml.sh:benchmark-vllm-sglang-and-tensorrt-llm-on-real-trafficMeasure offered load, completed streams, rejections and first-token latency. Use a real HTTP client and retain errors in the benchmark record.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/agent-engineering/bound-a-local-repair-process/aiml.sh:bound-a-local-repair-processRestrict a retry-condition patch, write each candidate into a fresh workspace and run independent behavioral checks in a separate process with a timeout.Sat, 24 Oct 2026 00:00:00 GMTAgent Design, Optimization and Harnesseshttps://aiml.sh/courses/llm-from-scratch/build-a-byte-pair-encoding-tokenizer/aiml.sh:build-a-byte-pair-encoding-tokenizerTrain a byte-pair encoding tokenizer on a small synthetic story corpus in plain Python, save it as a versioned file and test what it cannot encode.Sat, 24 Oct 2026 00:00:00 GMTBuild a Small Language Model From Scratchhttps://aiml.sh/courses/agent-engineering/build-a-coding-agent-loop/aiml.sh:build-a-coding-agent-loopBuild a coding-agent loop that streams model messages, runs repository tools, checks a real test repair and stops at a fixed turn limit.Sat, 24 Oct 2026 00:00:00 GMTAgent Design, Optimization and Harnesseshttps://aiml.sh/courses/context-and-memory/build-a-context-assembler/aiml.sh:build-a-context-assemblerAssemble instructions, tools, history and retrieved passages into one model request. Keep required sections whole and record optional content that does not fit.Sat, 24 Oct 2026 00:00:00 GMTContext and Memory Engineering for AI Agentshttps://aiml.sh/courses/context-and-memory/build-a-context-selection-benchmark/aiml.sh:build-a-context-selection-benchmarkCompare four history-selection strategies on shared cases, budgets and graders. Inspect retained facts, stale facts and strategy-specific failures.Sat, 24 Oct 2026 00:00:00 GMTContext and Memory Engineering for AI Agentshttps://aiml.sh/courses/building-ai-systems/build-a-guardrails-proxy-with-streaming-output-tests/aiml.sh:build-a-guardrails-proxy-with-streaming-output-testsBuild a streaming proxy that redacts or blocks email addresses and card numbers across chunk boundaries. Test it with saved streams and failures.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/llm-from-scratch/build-a-kv-cache-for-llm-generation/aiml.sh:build-a-kv-cache-for-llm-generationReuse attention keys and values during generation. Test cached and uncached output equivalence, and count the computation avoided by the cache.Sat, 24 Oct 2026 00:00:00 GMTBuild a Small Language Model From Scratchhttps://aiml.sh/courses/context-and-memory/build-a-memory-compaction-benchmark/aiml.sh:build-a-memory-compaction-benchmarkReplay a conversation with changed and forgotten facts. Compare four context strategies under one budget and inspect current, stale and deleted facts.Sat, 24 Oct 2026 00:00:00 GMTContext and Memory Engineering for AI Agentshttps://aiml.sh/courses/context-and-memory/build-a-memory-service/aiml.sh:build-a-memory-serviceStore assistant memory in SQLite across restarts. Scope each read by user and task, then test that another scope cannot retrieve the stored facts.Sat, 24 Oct 2026 00:00:00 GMTContext and Memory Engineering for AI Agentshttps://aiml.sh/courses/context-and-memory/build-a-memory-usefulness-eval/aiml.sh:build-a-memory-usefulness-evalMeasure needed facts, unnecessary memory, stale lines and forgotten content sent with requests. Compare usefulness with recall and storage growth.Sat, 24 Oct 2026 00:00:00 GMTContext and Memory Engineering for AI Agentshttps://aiml.sh/courses/building-ai-systems/build-a-model-selection-harness/aiml.sh:build-a-model-selection-harnessDefine requirements for a small task, test two local models on development and held-out cases, and inspect the recorded model-selection rule.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/llm-from-scratch/build-a-next-token-prediction-baseline/aiml.sh:build-a-next-token-prediction-baselineScore uniform, unigram and bigram models on held-out stories with cross-entropy, and record the prediction a model that sees one token back cannot make.Sat, 24 Oct 2026 00:00:00 GMTBuild a Small Language Model From Scratchhttps://aiml.sh/courses/building-ai-systems/build-a-paired-comparison-report/aiml.sh:build-a-paired-comparison-reportCompare prompt versions on paired cases and seeds. Bootstrap differences over cases, source groups and trials, then inspect uncertainty by slice.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/building-ai-systems/build-a-policy-boundary-for-untrusted-agent-handoffs-and-tool-requests/aiml.sh:build-a-policy-boundary-for-untrusted-agent-handoffs-and-tool-requestsCheck Docuctl proposals against trusted page and reviewer records. Keep handoffs untrusted and hold publication after external input arrives.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/building-ai-systems/build-a-portable-model-adapter-with-contract-tests/aiml.sh:build-a-portable-model-adapter-with-contract-testsUse one chat contract for Ollama’s native and OpenAI-compatible APIs. Test recorded traffic for request, response and error behavior.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/building-ai-systems/build-a-repeated-trial-benchmark/aiml.sh:build-a-repeated-trial-benchmarkRepeat evaluation cases across sampled model replies, log each trial and resume interrupted runs. Report per-case passes, pass@k and pass^k.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/llm-from-scratch/build-a-server-for-your-trained-llm/aiml.sh:build-a-server-for-your-trained-llmServe the trained model through a validated local HTTP interface. Compare served and direct generation, and evaluate on held-out story tasks.Sat, 24 Oct 2026 00:00:00 GMTBuild a Small Language Model From Scratchhttps://aiml.sh/courses/llm-from-scratch/build-a-small-language-model-training-pipeline/aiml.sh:build-a-small-language-model-training-pipelineTrain a one-block transformer on bundled stories using batches and Adam. Inspect loss and compare it with a bigram baseline on held-out stories.Sat, 24 Oct 2026 00:00:00 GMTBuild a Small Language Model From Scratchhttps://aiml.sh/courses/building-mlops/build-a-supervised-learning-baseline-with-reproducible-dataset-splits/aiml.sh:build-a-supervised-learning-baseline-with-reproducible-dataset-splitsTrain a rice-variety classifier with stable data splits and training-only preprocessing. Compare it with a majority rule and investigate changed results.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/llm-from-scratch/build-a-transformer-block-with-positional-information/aiml.sh:build-a-transformer-block-with-positional-informationCombine causal attention, normalization, skip paths and a feed-forward network. Add position embeddings and check the complete block’s gradients.Sat, 24 Oct 2026 00:00:00 GMTBuild a Small Language Model From Scratchhttps://aiml.sh/courses/building-ai-systems/build-authentication-and-authorization-boundaries-for-users-and-tools/aiml.sh:build-authentication-and-authorization-boundaries-for-users-and-toolsDelegate a Docuctl tool call with limited authority. Test role, tenant, actor, audience, expiry and replay checks without forwarding session tokens.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/llm-from-scratch/build-causal-self-attention-and-inspect-intermediate-outputs/aiml.sh:build-causal-self-attention-and-inspect-intermediate-outputsImplement causal self-attention and its backward pass in plain Python. Inspect intermediate tensors and test that future tokens remain hidden.Sat, 24 Oct 2026 00:00:00 GMTBuild a Small Language Model From Scratchhttps://aiml.sh/courses/llm-from-scratch/build-checkpoints-and-resumable-training/aiml.sh:build-checkpoints-and-resumable-trainingSave checkpoints bound to training inputs, select parameters by development loss and test that an interrupted run resumes to the same result.Sat, 24 Oct 2026 00:00:00 GMTBuild a Small Language Model From Scratchhttps://aiml.sh/courses/context-and-memory/build-context-compaction/aiml.sh:build-context-compactionReplace removed conversation turns with cited notes. Test retained, lost and stale facts instead of assuming a shorter context preserves every fact.Sat, 24 Oct 2026 00:00:00 GMTContext and Memory Engineering for AI Agentshttps://aiml.sh/courses/building-ai-systems/build-end-to-end-request-tracing-with-evidence-provenance/aiml.sh:build-end-to-end-request-tracing-with-evidence-provenancePropagate a trace across two local Docuctl processes. Inspect parent links and distinguish current, stale and dropped sources before returning a citation.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/context-and-memory/build-isolated-contexts/aiml.sh:build-isolated-contextsScope context to the requesting app, user, task or agent. Compare a leaking shared list with a scoped store and test restricted worker handoffs.Sat, 24 Oct 2026 00:00:00 GMTContext and Memory Engineering for AI Agentshttps://aiml.sh/courses/building-ai-systems/build-llm-monitoring-for-latency-cost-and-errors/aiml.sh:build-llm-monitoring-for-latency-cost-and-errorsAggregate simulated model-service requests into latency, token, cost, error and queue metrics. Test alert transitions and histogram estimates.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/context-and-memory/build-memory-consolidation-with-conflict-tests/aiml.sh:build-memory-consolidation-with-conflict-testsMerge duplicate memories and resolve conflicts using source and recency rules. Retain superseded facts with records of the consolidation decision.Sat, 24 Oct 2026 00:00:00 GMTContext and Memory Engineering for AI Agentshttps://aiml.sh/courses/context-and-memory/build-memory-provenance-and-deletion/aiml.sh:build-memory-provenance-and-deletionTrack memory sources, corrections and expiry in SQLite. Delete forgotten content while retaining a text-free record, and test scope isolation.Sat, 24 Oct 2026 00:00:00 GMTContext and Memory Engineering for AI Agentshttps://aiml.sh/courses/context-and-memory/build-memory-retrieval/aiml.sh:build-memory-retrievalRetrieve working notes, facts and past events with separate rules and token quotas. Compare this with a pooled ranking on the same memory tasks.Sat, 24 Oct 2026 00:00:00 GMTContext and Memory Engineering for AI Agentshttps://aiml.sh/courses/llm-from-scratch/build-one-training-step-by-hand/aiml.sh:build-one-training-step-by-handBuild a 20-parameter next-token model in plain Python, compute every gradient by hand, check them against finite differences and apply one update.Sat, 24 Oct 2026 00:00:00 GMTBuild a Small Language Model From Scratchhttps://aiml.sh/courses/building-mlops/build-point-in-time-feature-joins/aiml.sh:build-point-in-time-feature-joinsKeep original and corrected feature values. Reconstruct historical training rows using only the values available at each prediction time.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/llm-from-scratch/build-text-generation-with-sampling-controls/aiml.sh:build-text-generation-with-sampling-controlsGenerate text from the trained model with greedy choice, temperature and top-k. Define stop rules and compare settings on development openings.Sat, 24 Oct 2026 00:00:00 GMTBuild a Small Language Model From Scratchhttps://aiml.sh/courses/llm-inference/cache-shared-prompt-prefixes-and-route-to-warm-gpus/aiml.sh:cache-shared-prompt-prefixes-and-route-to-warm-gpusImplement radix-prefix matching, pinning and eviction. Compare cache affinity with queue waiting and inspect real reuse of model cache tensors.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/agent-engineering/calibrate-outcome-graders-and-compare-repeated-trials/aiml.sh:calibrate-outcome-graders-and-compare-repeated-trialsExpose false positives in a format-only grader, compare it with the task contract and calculate what any-success and repeated-success metrics actually mean.Sat, 24 Oct 2026 00:00:00 GMTAgent Design, Optimization and Harnesseshttps://aiml.sh/courses/building-mlops/capture-dependencies-and-configuration-for-an-existing-trainer/aiml.sh:capture-dependencies-and-configuration-for-an-existing-trainerCheck whether a declared 300-epoch recipe would actually run for 300 epochs. Capture the inputs and Python runtime without launching training.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/building-ai-systems/choose-an-ai-system-architecture/aiml.sh:choose-an-ai-system-architectureBuild a version-aware documentation lookup and an approval-bound action proposal. Use their contracts to choose an application boundary.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/llm-inference/choose-hardware-for-inference/aiml.sh:choose-the-right-gpu-for-prefill-and-decodeUse a roofline estimate to distinguish compute and bandwidth limits. Compare memory fit and explain why hardware choice depends on workload.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/llm-from-scratch/compare-a-handwritten-transformer-with-pytorch/aiml.sh:compare-a-handwritten-transformer-with-pytorchPort the existing one-block transformer to CPU tensors, then compare logits, loss and every parameter gradient with explicit attention and PyTorch SDPA.Sat, 24 Oct 2026 00:00:00 GMTBuild a Small Language Model From Scratchhttps://aiml.sh/courses/agent-engineering/compare-a-workflow-with-a-bounded-policy/aiml.sh:compare-a-workflow-with-a-bounded-policyCompare a keyword route, fixed search/read workflow and scripted decision loop on shared documentation tasks. Measure outcomes and tool work.Sat, 24 Oct 2026 00:00:00 GMTAgent Design, Optimization and Harnesseshttps://aiml.sh/courses/building-ai-systems/compare-full-and-low-rank-updates/aiml.sh:compare-full-and-low-rank-updatesTrain full and low-rank updates of a small text classifier, compare held-out predictions and count the parameters each method changes.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/building-ai-systems/compare-packed-weight-quantization-and-output-error/aiml.sh:compare-packed-weight-quantization-and-output-errorPack trained matrices into eight-bit values or paired four-bit codes. Compare numeric payload sizes and output errors using dequantized CPU inference.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/building-mlops/connect-prediction-errors-to-product-costs/aiml.sh:connect-prediction-errors-to-product-costsCompare two classifiers on the same validation cases and show how explicit costs change the consequences of their errors.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/building-ai-systems/create-an-evaluation-set-and-a-failure-taxonomy/aiml.sh:create-an-evaluation-set-and-a-failure-taxonomyTurn version, environment, role and action requirements into evaluation cases. Distinguish failed task contracts from incomplete execution.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/agent-engineering/curate-traces-before-selecting-an-improvement/aiml.sh:curate-traces-before-selecting-an-improvementDeduplicate versioned questions, keep task families within one split and save the actual decisions and tool observations for each cleaned case.Sat, 24 Oct 2026 00:00:00 GMTAgent Design, Optimization and Harnesseshttps://aiml.sh/courses/llm-inference/choose-a-latency-throughput-operating-point/aiml.sh:cut-per-token-latency-without-hurting-throughputChoose a feasible latency and throughput operating point. Inspect memory limits, communication assumptions and efficiency per serving device.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/llm-inference/profile-inference-host-overhead/aiml.sh:debug-a-gpu-stuck-at-30-percent-utilization-under-loadProfile host operators around an actual Metal model forward. Read recorded events without treating CPU timings as GPU kernel measurements.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/building-ai-systems/define-a-task-specific-answer-rubric/aiml.sh:define-a-task-specific-answer-rubricWrite source-based criteria for a documentation answer, bind example judgements to exact revisions and keep pending review separate from scored outcomes.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/agent-engineering/delegate-independent-tasks-with-explicit-worker-state/aiml.sh:delegate-independent-tasks-with-explicit-worker-stateFan out two documentation requests into separate bounded runtimes. Gather results by task ID and account for dispatch, merge and incomplete-worker behavior.Sat, 24 Oct 2026 00:00:00 GMTAgent Design, Optimization and Harnesseshttps://aiml.sh/courses/llm-inference/size-serving-capacity/aiml.sh:design-a-1m-user-inference-fleet-and-cut-cost-per-million-tokens-by-70-percentSize serving replicas from accepted workload-matched results. Convert request demand to peak capacity and include reserves and stated cost assumptions.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/building-mlops/design-temporal-and-entity-aware-evaluation-splits/aiml.sh:design-temporal-and-entity-aware-evaluation-splitsSplit repeated-user observations by user, time or both. Choose evaluation windows that match the population your predictor will serve.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/building-mlops/distinguish-schema-freshness-and-distribution-signals/aiml.sh:distinguish-schema-freshness-and-distribution-signalsValidate event fields and feature values before considering freshness. A local executable Rice rollout and monitoring continuation.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/llm-inference/enforce-multi-tenant-fairness-priority-queues-and-admission-limits/aiml.sh:enforce-multi-tenant-fairness-priority-queues-and-admission-limitsImplement tenant queues, token budgets and deficit round-robin scheduling. Inspect admission limits and the latency cost of protecting interactive work.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/building-mlops/export-a-full-precision-inference-bundle/aiml.sh:export-a-full-precision-inference-bundleKeep the trained scaler and classifier together without rounding serving parameters. A local Rice-classifier lab with explicit failure checks.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/llm-inference/fix-long-prompt-stalls-with-chunked-prefill/aiml.sh:fix-long-prompt-stalls-with-chunked-prefillSchedule prompt chunks around active decode work. Compare modeled token budgets with a real cached-forward experiment and explain added overhead.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/llm-inference/trace-an-inference-request/aiml.sh:follow-an-inference-request-from-prompt-to-first-tokenTrace tokenization, prefill and decode with a pinned open model. Compare actual cached and uncached token intervals from a local inference run.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/building-mlops/gate-a-model-candidate-on-validation-results/aiml.sh:gate-a-model-candidate-on-validation-resultsMake a quality decision about the exact exported candidate using explicit count-based checks. A local Rice-classifier lab with explicit failure checks.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/building-ai-systems/inspect-a-direct-preference-optimization-update/aiml.sh:inspect-a-direct-preference-optimization-updateAlign completion-token scores with their preceding prompt positions. Execute a focused local CPU experiment and inspect its raw result.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/building-mlops/join-delayed-outcomes-and-review-retraining/aiml.sh:join-delayed-outcomes-and-review-retrainingJoin finalized labels to predictions by identity rather than list position. A local executable Rice rollout and monitoring continuation.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/building-ai-systems/measure-kv-cache-work/aiml.sh:measure-kv-cache-workImplement one causal attention layer, reuse previous keys and values, and compare equivalent outputs, counted work and local CPU timings.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/building-ai-systems/measure-quality-by-request-slice-and-asymmetric-failure-cost/aiml.sh:measure-quality-by-request-slice-and-asymmetric-failure-costInspect request-group coverage and explicit error costs. Keep critical failures, incomplete runs and empty groups visible beside the overall score.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/llm-inference/compare-eager-and-compiled-normalization/aiml.sh:profile-a-slow-model-and-fuse-a-bottleneck-kernel-in-tritonCompare eager and compiled normalization on real Metal tensors. Check numerical agreement and synchronized timings across several input shapes.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/llm-inference/evaluate-model-quantization/aiml.sh:quantize-a-model-and-measure-the-quality-dropConvert real model linear layers to INT8 and compare development likelihood. Inspect the observed regression before accepting a serving candidate.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/building-mlops/record-data-availability-and-label-delay/aiml.sh:record-data-availability-and-label-delayReconstruct the inputs available for each prediction. Account for late events and wait for outcome labels before assembling training rows.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/building-mlops/register-and-select-a-content-addressed-model/aiml.sh:register-and-select-a-content-addressed-modelBind a local model selection to separately retained expected bytes. A local Rice-classifier lab with explicit failure checks.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/llm-inference/replace-static-batching-with-continuous-batching/aiml.sh:replace-static-batching-with-continuous-batchingCompare fixed and changing batch membership with a request scheduler. Inspect actual local model interleaving and its latency tradeoffs.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/building-mlops/roll-back-a-model-with-feature-contract-checks/aiml.sh:roll-back-a-model-with-feature-contract-checksSelect a prior model by its retained digest, refuse incompatible feature contracts, and verify the local serving selection over actual HTTP.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/building-mlops/route-a-sticky-canary-and-compare-cohort-outcomes/aiml.sh:route-a-sticky-canary-and-compare-cohort-outcomesAssign repeated caller IDs to a stable canary cohort with an explicit salt and allocation. A local executable Rice rollout and monitoring continuation.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/agent-engineering/run-a-search-and-read-decision-loop/aiml.sh:run-a-search-and-read-decision-loopSeparate policy decisions from tool execution. Enforce action schemas, tool and turn limits, and citation provenance before checking the returned command.Sat, 24 Oct 2026 00:00:00 GMTAgent Design, Optimization and Harnesseshttps://aiml.sh/courses/building-ai-systems/run-group-relative-policy-updates-on-sampled-completions/aiml.sh:run-group-relative-policy-updates-on-sampled-completionsSample completions from a tiny CPU policy, compute group-relative rewards, and inspect clipped updates with fixed old scores and a frozen reference.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/building-ai-systems/separate-source-support-action-correctness-and-format-validity/aiml.sh:separate-source-support-action-correctness-and-format-validityDiagnose a structured answer without letting valid JSON, a matching citation or a correct command stand in for the other requirements.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/building-mlops/serve-a-versioned-predictor-through-batch-and-http/aiml.sh:serve-a-versioned-predictor-through-batch-and-httpReuse one validated preprocessing and prediction function across two adapters. A local Rice-classifier lab with explicit failure checks.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/building-ai-systems/set-task-acceptance-criteria-and-request-resource-budgets/aiml.sh:set-task-acceptance-criteria-and-request-resource-budgetsDefine task acceptance and request resource limits. Reuse application graders and distinguish admission decisions from post-execution checks.Sat, 24 Oct 2026 00:00:00 GMTBuilding AI Systemshttps://aiml.sh/courses/building-mlops/shadow-predictions-without-changing-live-responses/aiml.sh:shadow-predictions-without-changing-live-responsesRoute a request to the live model while recording a separate candidate prediction. A local executable Rice rollout and monitoring continuation.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/llm-inference/size-gpu-memory-for-model-weights-and-the-kv-cache/aiml.sh:size-gpu-memory-for-model-weights-and-the-kv-cacheCalculate weight and attention-cache payloads with consistent byte units. Compare grouped-query cache sizes and budget runtime memory separately.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/llm-inference/verify-tiled-attention/aiml.sh:speed-up-long-contexts-with-flashattentionImplement tiled attention with online softmax on the CPU. Verify numerical agreement and distinguish storage savings from measured GPU runtime.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/llm-inference/split-prefill-and-decode-pools-and-serve-large-moe-models/aiml.sh:split-prefill-and-decode-pools-and-serve-large-moe-modelsCalculate cache-transfer payloads before separating prefill and decode. Explore bandwidth and expert-placement assumptions without claiming fleet benchmarks.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/llm-inference/stop-memory-waste-with-paged-kv-caching/aiml.sh:stop-memory-waste-with-paged-kv-cachingImplement a paged cache allocator with reference counts and copy-on-write. Inspect ownership and fragmentation using concrete token markers.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/llm-inference/stream-tokens-over-sse-and-cancel-dropped-requests/aiml.sh:stream-tokens-over-sse-and-cancel-dropped-requestsBuild a local streaming HTTP server with bounded admission and disconnect cleanup. Test SSE framing and iterator release over real sockets.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/building-mlops/test-feature-freshness-and-training-serving-parity/aiml.sh:test-feature-freshness-and-training-serving-parityCheck whether a feature is available and recent enough, then compare historical and current reads using identical inputs.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/llm-inference/investigate-inference-tail-latency/aiml.sh:track-inference-metrics-and-fix-10x-p99-latency-spikesAggregate request records to investigate tail latency. Keep timing boundaries, preemptions and admission outcomes distinct when testing hypotheses.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/building-mlops/track-run-identity-and-select-an-immutable-training-artifact/aiml.sh:track-run-identity-and-select-an-immutable-training-artifactInspect a completed training run and verify its retained inputs, logs and results. Select its exact evaluation report without rerunning the trainer.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/agent-engineering/tune-retrieval-policy-on-separated-task-families/aiml.sh:tune-retrieval-policy-on-separated-task-familiesChoose alias expansion and candidate count on development tasks. Freeze the selected configuration before measuring source recall on separate task families.Sat, 24 Oct 2026 00:00:00 GMTAgent Design, Optimization and Harnesseshttps://aiml.sh/courses/llm-inference/use-speculative-decoding-and-know-when-it-hurts/aiml.sh:use-speculative-decoding-and-know-when-it-hurtsImplement speculative token acceptance and residual resampling. Verify the target distribution and separate correctness from assumed speedups.Sat, 24 Oct 2026 00:00:00 GMTLLM Inference Engineeringhttps://aiml.sh/courses/building-mlops/validate-a-dataset-and-quarantine-invalid-records/aiml.sh:validate-a-dataset-and-quarantine-invalid-recordsValidate an event batch, preserve rejected records and block the batch when parsing or schema checks fail. Inspect the rejection reasons.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/agent-engineering/verify-edits-independently-of-completion-claims/aiml.sh:verify-edits-independently-of-completion-claimsRun controlled repair candidates through the existing harness. Preserve completion claims as observations while grading the behavior of each fresh candidate.Sat, 24 Oct 2026 00:00:00 GMTAgent Design, Optimization and Harnesseshttps://aiml.sh/courses/building-mlops/version-data-schemas-and-feature-definitions/aiml.sh:version-data-schemas-and-feature-definitionsTie dataset, schema and feature versions to their contents. Detect silent changes before computing feature values from the selected manifest.Sat, 24 Oct 2026 00:00:00 GMTBuilding MLOpshttps://aiml.sh/courses/agentic-rag/add-search-and-read-tools/aiml.sh:add-search-and-read-toolsExpose bounded documentation search and passage reads, bind handles to a pinned corpus and exact request, and inspect provenance before accepting tool output.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/add-vector-retrieval/aiml.sh:add-vector-retrievalQuery an actual LangChain vector store with preserved Qwen embeddings, refuse unrecorded inputs, and inspect scores without new model calls.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/apply-filters-across-retrievers/aiml.sh:apply-filters-across-retrieversApply one version policy to SQLite lexical and LangChain vector retrieval before top-k, and compare eligible candidates and post-filter underfill.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/assemble-context-within-a-token-budget/aiml.sh:assemble-context-within-a-token-budgetFit validated RAG windows into a complete templated request using the pinned local tokenizer, with whole-window admission, output reserve and failure records.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/authorize-collection-searches/aiml.sh:authorize-collection-searchesResolve synthetic fixture identities, filter granted collections before top-k, and recheck issued passage handles before returning source text.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/build-a-rag-eval-dataset/aiml.sh:build-a-rag-eval-datasetCreate eight source-bound developer questions, label relevant passages by documentation version, and validate answerability and frozen held-out targets.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/build-a-retrieval-inspector/aiml.sh:build-a-retrieval-inspectorBuild a same-origin retrieval inspector, verify response bytes and request identity before rendering, and inspect real loading, empty, limited and error states.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/build-developer-knowledge-assistant/aiml.sh:build-developer-knowledge-assistantBuild a local docs assistant with a browser UI, CLI, versioned retrieval, collection permissions, cited answers and application evaluation.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/build-hybrid-retrieval/aiml.sh:build-hybrid-retrievalCombine native BM25 and preserved-vector retrieval with reciprocal rank fusion, inspect each contribution, and retain filtering and failure records.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/cache-retrieval-results/aiml.sh:cache-retrieval-resultsCache completed retrieval sources with principal, policy, corpus and normalized-request binding, bounded TTL/LRU retention and explicit failure bypass.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/choose-chunk-boundaries/aiml.sh:choose-chunk-boundariesCompare fixed-character, paragraph and heading-aware chunk boundaries on a synthetic developer document, with exact source spans and stable identity checks.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/compare-agents-and-pipelines/aiml.sh:compare-agents-and-pipelinesCompare fixed, scripted and Gemini-selected retrieval policies on frozen questions, enforce tool limits, and retain invalid proposals and incomplete outcomes.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/compare-retrieval-configurations/aiml.sh:compare-retrieval-configurationsCompare frozen lexical, vector, fused and coverage-reranked configurations on preserved development questions, with exact target mapping and retained misses.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/connect-retrieval-tools-through-mcp/aiml.sh:connect-retrieval-tools-through-mcpExpose pinned documentation tools through an actual MCP stdio server, initialize an SDK client, preserve issued read handles and inspect real tool failures.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agents-cli/create-and-run-an-agent-project-with-agents-cli/aiml.sh:create-and-run-an-agent-project-with-agents-cliScaffold a prototype ADK agent with agents-cli, inspect agents-cli-manifest.yaml, run agents-cli lint, and serve multi-turn ADK and A2A sessions locally.Fri, 23 Oct 2026 00:00:00 GMTBuilding, Evaluating, and Deploying Agents with Google Agents CLIhttps://aiml.sh/courses/agents-cli/evaluate-an-agent-with-agents-cli-eval/aiml.sh:evaluate-an-agent-with-agents-cli-evalAuthor single-turn and continued-conversation datasets, generate ADK traces with agents-cli eval generate, and score custom metrics in eval grade.Fri, 23 Oct 2026 00:00:00 GMTBuilding, Evaluating, and Deploying Agents with Google Agents CLIhttps://aiml.sh/courses/agentic-rag/evaluate-retrieval-agent-tasks/aiml.sh:evaluate-retrieval-agent-tasksEvaluate deterministic retrieval policies with separate execution and source checks, exact target joins, repeated trials and an explicit fixture release gate.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/follow-a-rag-request-end-to-end/aiml.sh:follow-a-rag-request-end-to-endFollow a question through document search, context assembly and the model request, then check the answer’s source links.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/handle-conflicting-documentation/aiml.sh:handle-conflicting-documentationFilter synthetic developer documentation by the requested version, compare explicit contract directives, and withhold answer requests when claims conflict.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/handle-ingestion-failures/aiml.sh:handle-ingestion-failuresKeep accepted documents and explicit failure records together, and return an incomplete batch status when a source cannot be loaded.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/how-bm25-ranks-documents/aiml.sh:how-bm25-ranks-documentsInspect native SQLite FTS5 BM25 scores, indexed terms and document lengths, and keep matching passages separate from answer validation.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/how-rag-works/aiml.sh:how-rag-worksLearn how an application searches documents, supplies relevant passages to a language model and links its answer to sources.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/limit-concurrent-rag-requests/aiml.sh:limit-concurrent-rag-requestsBound active retrieval and waiting work in one process; inspect actual FIFO promotion, queue refusals, wait deadlines and cancellation-safe lease cleanup.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/limit-retrieval-agent-runs/aiml.sh:limit-retrieval-agent-runsBound a scripted retrieval run with action and source-byte admission checks, retain denied and failed attempts, and separate limits from task completion.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/load-developer-documentation/aiml.sh:load-developer-documentationLoad a pinned local UTF-8 documentation snapshot into records that keep the source identity, version, URL and text together.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agents-cli/provision-terraform-and-deploy-with-agents-cli/aiml.sh:provision-terraform-and-deploy-with-agents-cliSeparate Terraform infrastructure from code rollouts, validate Agent Runtime and Cloud Run rollouts with agents-cli deploy --dry-run, and inspect metadata.Fri, 23 Oct 2026 00:00:00 GMTBuilding, Evaluating, and Deploying Agents with Google Agents CLIhttps://aiml.sh/courses/agents-cli/publish-to-gemini-and-configure-agent-observability/aiml.sh:publish-to-gemini-and-configure-agent-observabilityRegister a deployed agent with Gemini Enterprise in ADK or A2A mode using agents-cli publish gemini, and query two-tier GCS and BigQuery telemetry.Fri, 23 Oct 2026 00:00:00 GMTBuilding, Evaluating, and Deploying Agents with Google Agents CLIhttps://aiml.sh/courses/agentic-rag/record-a-rag-baseline/aiml.sh:record-a-rag-baselineRecord a fixed RAG pipeline on eight versioned questions, separate retrieval from answer checks, and preserve missing answers without retries.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/refine-queries-and-search-again/aiml.sh:refine-queries-and-search-againRun a scripted retrieval controller that observes source passages, revises a query under an explicit rule, and retains its attempts and stopping reasons.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/release-the-rag-application/aiml.sh:release-the-rag-applicationStage the existing retrieval service locally, bind its files to a retained release identity, block failed task checks, and test artifact-pointer rollback.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/rerank-retrieved-passages/aiml.sh:rerank-retrieved-passagesCompare coverage, Gemini relevance and a pinned cross-encoder on frozen candidates, preserving passage identity and honest ranking limits.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/resume-document-ingestion/aiml.sh:resume-document-ingestionResume a pinned SQLite ingestion job after an injected interruption, checking committed sources, chunk rows and atomic completion checkpoints.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/retrieve-surrounding-context/aiml.sh:retrieve-surrounding-contextExpand explicit seed chunks into bounded source-local context windows, merge overlaps, and preserve section, revision and seed provenance before assembly.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/run-rag-evals-on-an-updated-corpus/aiml.sh:run-rag-evals-on-an-updated-corpusCompare fixed development retrieval cases across a synthetic corpus update, refuse stale target hashes, and retain deleted targets outside paired means.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/serve-the-rag-application-with-fastapi/aiml.sh:serve-the-rag-application-with-fastapiServe the existing retrieval core through FastAPI, validate bounded requests, preserve source metadata and response hashes, and inspect local HTTP failures.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/stream-rag-answers/aiml.sh:stream-rag-answersCompare preserved answer replay with actual Gemini SSE events, keep partial output provisional, and validate completion and context-bound source IDs.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agents-cli/synthesize-user-simulations-and-compare-eval-runs/aiml.sh:synthesize-user-simulations-and-compare-eval-runsGenerate multi-turn user-simulated traces with agents-cli eval dataset synthesize, cluster losses with eval analyze, and diff runs with eval compare.Fri, 23 Oct 2026 00:00:00 GMTBuilding, Evaluating, and Deploying Agents with Google Agents CLIhttps://aiml.sh/courses/agentic-rag/trace-rag-requests/aiml.sh:trace-rag-requestsTrace actual local retrieval executions with isolated context IDs, fixed-field redacted spans, bounded retention and reproducible failure diagnostics.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/track-passage-sources/aiml.sh:track-passage-sourcesBind an exact passage to its source path, byte hash and line span, then reject stale or altered records before displaying the text.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agentic-rag/update-and-delete-indexed-content/aiml.sh:update-and-delete-indexed-contentUpdate or delete source and chunk rows with a SQLite FTS5 index in one transaction, then check actual search results and rollback behavior.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/agents-cli/upgrade-and-extend-an-agents-cli-project/aiml.sh:upgrade-and-extend-an-agents-cli-projectAdd Cloud Run and CI/CD scaffolding in place with scaffold enhance, run 3-way template merges with scaffold upgrade, and wire agents-cli-extension.yaml.Fri, 23 Oct 2026 00:00:00 GMTBuilding, Evaluating, and Deploying Agents with Google Agents CLIhttps://aiml.sh/courses/agentic-rag/version-documents-and-chunks/aiml.sh:version-documents-and-chunksCreate source and chunk revisions from pinned documentation, compare text and metadata changes, and reject inconsistent saved snapshots.Fri, 23 Oct 2026 00:00:00 GMTBuilding Agentic RAG Applicationshttps://aiml.sh/courses/rag/build-a-rag-pipeline-with-citations/aiml.sh:build-a-rag-pipeline-with-citationsRetrieve internal developer documentation with PGVectorStore, ask Gemini for source IDs and quotes, and reject citations outside the supplied chunks.Thu, 22 Oct 2026 00:00:00 GMTRetrieval-Augmented Generation (RAG)https://aiml.sh/courses/llm-evaluation/build-a-release-manifest-for-application-model-tokenizer-prompt-data-and-index-versions/aiml.sh:build-a-release-manifest-for-application-model-tokenizer-prompt-data-and-index-versionsHash a tracker docs-assistant bundle, compare its settings with a recorded eval run, and reject changed files or missing evaluated cases.Thu, 22 Oct 2026 00:00:00 GMTLLM Evalshttps://aiml.sh/courses/langgraph/add-a-long-term-memory-store/aiml.sh:add-a-long-term-memory-storeGive a LangGraph code-review agent memory across threads with a store keyed by repository and topic, and see the stale-rule and namespace-prefix failures.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with LangGraphhttps://aiml.sh/courses/langgraph/add-human-in-the-loop-interrupts/aiml.sh:add-human-in-the-loop-interruptsPause a LangGraph code-review agent with interrupt() before it posts a comment, then resume it with Command(resume=...) to approve, edit or reject.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with LangGraphhttps://aiml.sh/courses/google-adk-production/apply-safety-and-security-guardrails-to-agent-tools/aiml.sh:apply-safety-and-security-guardrails-to-agent-toolsGuard docs-assistant write tools with role checks, confirmation and path validation, then test refused calls and a followed prompt injection.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-live/build-a-custom-websocket-server-for-a-live-agent/aiml.sh:build-a-custom-websocket-server-for-a-live-agentBridge a WebSocket client to an ADK live agent with binary audio, JSON events, bounded concurrent sessions and cleanup when either direction stops.Wed, 21 Oct 2026 00:00:00 GMTBuilding Live and Voice Agents with Google ADK 2.0https://aiml.sh/courses/langgraph/build-a-graph-agent-with-tools/aiml.sh:build-a-graph-agent-with-toolsBuild a code-review agent as a LangGraph StateGraph with a ToolNode and tools_condition, run it on Gemini, and handle tool errors and the step limit.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with LangGraphhttps://aiml.sh/courses/google-adk/build-a-graph-workflow-from-nodes-and-edges/aiml.sh:build-a-graph-workflow-from-nodes-and-edgesBuild a release-notes workflow with function nodes, a typed agent node and a deterministic check that every commit reaches the output.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/build-an-agent-as-yaml-config-and-in-visual-builder/aiml.sh:build-an-agent-as-yaml-config-and-in-visual-builderDefine a Google ADK 2.0 agent in root_agent.yaml, load it with from_config, and test which YAML the loader and the Visual Builder save endpoint refuse.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk/build-an-llm-agent/aiml.sh:build-an-llm-agentBuild a documentation assistant with Google ADK, understand its model and instructions, and run it from the terminal or development Web UI.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/call-a-rest-api-from-an-openapi-spec-with-authentication/aiml.sh:call-a-rest-api-from-an-openapi-spec-with-authenticationTurn a release API's OpenAPI spec into ADK tools, send its API key outside the prompt, keep the agent read-only, and debug wrong keys.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-live/call-tools-during-a-live-conversation/aiml.sh:call-tools-during-a-live-conversationRun docs lookup and background tools in an ADK live conversation, inspect function responses, and reproduce tool errors and model-specific limits.Wed, 21 Oct 2026 00:00:00 GMTBuilding Live and Voice Agents with Google ADK 2.0https://aiml.sh/courses/google-adk/choose-an-adk-workflow-pattern/aiml.sh:choose-an-adk-workflow-patternCompare ADK workflow patterns by dependencies, control and failure policy, then inspect the real lesson graphs without calling a model.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/google-adk/connect-an-agent-to-mcp-servers-with-mcptoolset/aiml.sh:connect-an-agent-to-mcp-servers-with-mcptoolsetAttach a typed MCP server to a Google ADK 2.0 agent with McpToolset, expose only the tools it needs, read what reaches the model and catch a failed start.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/google-adk/delegate-to-sub-agents-in-a-collaborative-workflow/aiml.sh:delegate-to-sub-agents-in-a-collaborative-workflowBuild a documentation review team with a coordinator, two typed specialists and an event check that both delegations ran.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/deploy-an-adk-agent-to-agent-runtime/aiml.sh:deploy-an-adk-agent-to-agent-runtimeDeploy an ADK agent with the native CLI, call Agent Runtime session and streaming methods, and distinguish endpoint mistakes from user-session errors.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/deploy-an-adk-agent-to-cloud-run/aiml.sh:deploy-an-adk-agent-to-cloud-runServe an ADK agent on private Cloud Run, separate caller and runtime permissions, and inspect what happens to local sessions across a revision change.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk/evaluate-an-agents-tool-trajectory-and-response/aiml.sh:evaluate-an-agents-tool-trajectory-and-responseWrite evaluation cases for tool use and complete answers, run ADK AgentEvaluator and understand the limits of trajectory, ROUGE and judge scores.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-integrations/expose-an-agent-over-a2a-and-consume-it/aiml.sh:expose-an-agent-over-a2a-and-consume-itServe an ADK agent over A2A with to_a2a, call it from another agent with RemoteA2aAgent, and debug card, connection and timeout failures.Wed, 21 Oct 2026 00:00:00 GMTConnecting Agents to Tools and Services with Google ADK 2.0https://aiml.sh/courses/google-adk-production/expose-the-docs-assistant-as-an-mcp-server/aiml.sh:expose-the-docs-assistant-as-an-mcp-serverServe the docs assistant as one MCP tool with to_mcp_server, call it from the MCP SDK and another ADK agent, and fix key and timeout failures.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk/generate-and-review/aiml.sh:generate-and-reviewBuild an ADK writer and reviewer that compare a release announcement with changelog facts, then enforce the typed verdict in code.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-integrations/give-an-agent-skills-from-a-folder/aiml.sh:give-an-agent-skills-from-a-folderLoad a folder of Agent Skills into an ADK agent with SkillToolset, run skill scripts, and catch the failures that silently skip a skill's steps.Wed, 21 Oct 2026 00:00:00 GMTConnecting Agents to Tools and Services with Google ADK 2.0https://aiml.sh/courses/google-adk-live/guard-and-evaluate-a-voice-agent/aiml.sh:guard-and-evaluate-a-voice-agentTest ADK live input, output and tool callbacks, inspect audio delivered before a block, and check what a recorded speech evaluation really scored.Wed, 21 Oct 2026 00:00:00 GMTBuilding Live and Voice Agents with Google ADK 2.0https://aiml.sh/courses/google-adk/guard-model-and-tool-calls-with-callbacks-and-plugins/aiml.sh:guard-model-and-tool-calls-with-callbacks-and-pluginsGuard a Google ADK 2.0 agent with a plugin that strips AWS key IDs before Gemini sees them and a callback that keeps a file tool inside its workspace.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/google-adk/hierarchical-task-decomposition/aiml.sh:hierarchical-task-decompositionBuild a depth-limited ADK AgentTool tree for an upgrade guide, check its source paths and share a model-call budget across nested runs.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/mcp/implement-an-mcp-server-with-typed-tool-contracts/aiml.sh:implement-an-mcp-server-with-typed-tool-contractsImplement a task-tracker server with the MCP Python SDK whose typed tool schemas reject bad arguments and keep tool errors apart from protocol errors.Wed, 21 Oct 2026 00:00:00 GMTModel Context Protocol (MCP)https://aiml.sh/courses/google-adk-production/inject-tool-failures-with-environment-simulation/aiml.sh:inject-tool-failures-with-environment-simulationInject docs-service failures with ADK environment simulation, replay the recorded responses, and test the limits of retry guidance and error rules.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/mcp/integrate-mcp-tools-into-adk-and-langgraph-agents/aiml.sh:integrate-mcp-tools-into-adk-and-langgraph-agentsConnect one MCP server to a Google ADK agent with McpToolset and a LangGraph agent with langchain-mcp-adapters, and compare how each handles tool failures.Wed, 21 Oct 2026 00:00:00 GMTModel Context Protocol (MCP)https://aiml.sh/courses/google-adk/iterative-refinement/aiml.sh:iterative-refinementRefine a changelog line in an ADK 2.0 graph loop: a lint checks each draft, a back-edge sends problems to a reviser, and a state counter stops the loop.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/keep-long-sessions-in-budget-with-compaction-and-context-caching/aiml.sh:keep-long-sessions-in-budget-with-compaction-and-context-cachingCache the docs assistant's manual with ContextCacheConfig, summarize old turns with EventsCompactionConfig, and measure both in Gemini usage metadata.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/make-tools-parallel-ready/aiml.sh:make-tools-parallel-readyMeasure how ADK runs several function calls from one Gemini response, and make them overlap with async tools or a tool thread pool.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk/parallel-fan-out-and-gather/aiml.sh:parallel-fan-out-and-gatherRun ADK 2.0 workflow branches in parallel and merge them with a JoinNode, then see one failed branch fail the graph and a dynamic parent keep the rest.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/google-adk/pause-a-workflow-for-human-approval/aiml.sh:pause-a-workflow-for-human-approvalPause a migration workflow with the SQL attached, validate a reviewer decision and resume the same session to apply or reject the change.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/google-adk/persist-session-state-and-long-term-memory/aiml.sh:persist-session-state-and-long-term-memoryPersist Google ADK 2.0 sessions in SQLite with DatabaseSessionService, keep user state across restarts and recall earlier conversations with load_memory.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-integrations/query-bigquery-with-bigquerytoolset/aiml.sh:query-bigquery-with-bigquerytoolsetQuery docs analytics through ADK BigQueryToolset with SELECT-only execution, a per-query bytes cap, and offline tests for refused jobs.Wed, 21 Oct 2026 00:00:00 GMTConnecting Agents to Tools and Services with Google ADK 2.0https://aiml.sh/courses/google-adk-integrations/query-postgresql-through-mcp-toolbox-for-databases/aiml.sh:query-postgresql-through-mcp-toolbox-for-databasesConnect an ADK agent to PostgreSQL through MCP Toolbox with fixed SQL tools and a read-only role, and catch toolsets that fail to load.Wed, 21 Oct 2026 00:00:00 GMTConnecting Agents to Tools and Services with Google ADK 2.0https://aiml.sh/courses/google-adk-live/read-live-events-and-manage-live-sessions/aiml.sh:read-live-events-and-manage-live-sessionsFold ADK live events into captions, handle interruptions, inspect saved session history, and test reconnect behavior with and without a handle.Wed, 21 Oct 2026 00:00:00 GMTBuilding Live and Voice Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/resume-an-interrupted-agent-run/aiml.sh:resume-an-interrupted-agent-runMake a Google ADK 2.0 app resumable, finish a run whose process died inside a tool, and keep the retried tool from filing a duplicate issue.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/rewind-a-session-and-migrate-its-database/aiml.sh:rewind-a-session-and-migrate-its-databaseUndo turns in a Google ADK 2.0 session with rewind_async, see what it restores, and migrate a legacy pickle session database with adk migrate session.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk/route-a-workflow-with-conditional-edges/aiml.sh:route-a-workflow-with-conditional-edgesRoute documentation build logs with Python, use a default edge for incomplete runs and ask an agent for link suggestions only when needed.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/run-an-agent-from-event-triggers/aiml.sh:run-an-agent-from-event-triggersStart an ADK agent from Pub/Sub and Eventarc request bodies, authenticate local trigger calls, and deduplicate the review tool’s writes.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk/run-an-agent-from-python-and-read-its-event-stream/aiml.sh:run-an-agent-from-python-and-read-its-event-streamUse Python Runner to manage an ADK conversation, inspect model and tool events, select the final answer and handle streaming and model-call limits.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-live/run-live-agents-inside-a-workflow/aiml.sh:run-live-agents-inside-a-workflowSequence a voice intake and docs-answer agent with ADK Workflow, inspect the typed handoff, and test live modes and incomplete transcripts.Wed, 21 Oct 2026 00:00:00 GMTBuilding Live and Voice Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/save-and-load-files-with-artifacts/aiml.sh:save-and-load-files-with-artifactsStore an uploaded ship.toml and versioned reviews as ADK artifacts on disk, load files only when needed, and see which survive a restart.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/scaffold-and-deploy-an-agent-with-the-agents-cli/aiml.sh:scaffold-and-deploy-an-agent-with-the-agents-cliUse an Agents CLI project for an existing ADK agent, inspect deployment defaults, set explicit resource limits, and resume a recorded remote session.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-integrations/search-private-data-stores-with-agent-search/aiml.sh:search-private-data-stores-with-agent-searchSearch engineering documents with ADK VertexAiSearchTool, validate a team filter from application state, and inspect recorded grounding metadata.Wed, 21 Oct 2026 00:00:00 GMTConnecting Agents to Tools and Services with Google ADK 2.0https://aiml.sh/courses/google-adk-live/send-audio-images-and-video-and-configure-voice/aiml.sh:send-audio-images-and-video-and-configure-voiceSend PCM audio and JPEG frames to an ADK live agent, choose its voice, use push-to-talk signals, and diagnose format and configuration failures.Wed, 21 Oct 2026 00:00:00 GMTBuilding Live and Voice Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/serve-an-agent-over-http-and-drive-it-from-the-cli/aiml.sh:serve-an-agent-over-http-and-drive-it-from-the-cliServe a Google ADK 2.0 agent with adk api_server, call /run and /run_sse from a Python client, and script adk run with replay files and JSONL output.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/mcp/serve-mcp-over-stdio-and-streamable-http-and-test-the-contracts/aiml.sh:serve-mcp-over-stdio-and-streamable-http-and-test-the-contractsServe one MCP Python SDK server over stdio and Streamable HTTP, read the JSON-RPC requests the client sends, and pin its tool contracts with pytest.Wed, 21 Oct 2026 00:00:00 GMTModel Context Protocol (MCP)https://aiml.sh/courses/google-adk-live/stream-a-voice-conversation-with-run-live/aiml.sh:stream-a-voice-conversation-with-run-liveOpen an ADK live connection, send text and microphone audio, inspect transcripts and audio events, and compare saved answers with and without the docs page.Wed, 21 Oct 2026 00:00:00 GMTBuilding Live and Voice Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/test-an-agent-against-simulated-users/aiml.sh:test-an-agent-against-simulated-usersEvaluate a docs assistant with simulated novice and expert users, inspect multi-turn scores, and diagnose turn caps and unsupported metrics.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/trace-and-measure-agent-runs-with-opentelemetry/aiml.sh:trace-and-measure-agent-runs-with-opentelemetryRecord ADK spans and GenAI metrics with OpenTelemetry, inspect one docs-assistant run, and control prompt and tool content in exported spans.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/tune-a-run-with-runconfig-and-read-the-event-loop/aiml.sh:tune-a-run-with-runconfig-and-read-the-event-loopSet Google ADK 2.0 RunConfig per run to tag events, cap model calls and history, and time out the model, then trace when the Runner commits state.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-production/use-agent-platform-models-sessions-and-memory-bank/aiml.sh:use-agent-platform-models-sessions-and-memory-bankConnect a local ADK runner to managed sessions and Memory Bank, separate model and storage locations, and inspect one cross-session recall run.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk-integrations/use-gemini-built-in-tools-google-search-and-code-execution/aiml.sh:use-gemini-built-in-tools-google-search-and-code-executionAdd Gemini's Google Search and code execution to an ADK agent beside a function tool, cite answers from grounding metadata, and handle request errors.Wed, 21 Oct 2026 00:00:00 GMTConnecting Agents to Tools and Services with Google ADK 2.0https://aiml.sh/courses/google-adk-production/write-a-custom-eval-metric-and-optimize-the-instruction/aiml.sh:write-a-custom-eval-metric-and-optimize-the-instructionScore citations against completed docs reads, optimize an ADK instruction with GEPA, and inspect copied facts and limits of the validation set.Wed, 21 Oct 2026 00:00:00 GMTRunning and Deploying Agents with Google ADK 2.0https://aiml.sh/courses/google-adk/write-dynamic-workflows-in-python-code/aiml.sh:write-dynamic-workflows-in-python-codeReview changed documentation pages with dynamic child nodes, deterministic lint checks and a bounded agent fix loop with retries and timeouts.Wed, 21 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/langgraph/add-branches-and-stop-conditions/aiml.sh:build-a-bounded-agent-state-machine-with-branches-and-stop-conditionsReplace tools_condition with a custom LangGraph router that stops repeated tool calls and runs near recursion_limit, and test every branch without a model.Tue, 20 Oct 2026 00:00:00 GMTBuilding Agents with LangGraphhttps://aiml.sh/courses/google-adk/add-typed-function-tools/aiml.sh:build-a-function-calling-router-with-typed-tool-contractsGive a Google ADK 2.0 agent typed function tools, validate every call with Pydantic before it runs, and require human approval for the tool that writes.Tue, 20 Oct 2026 00:00:00 GMTBuilding Agents with Google ADK 2.0https://aiml.sh/courses/langgraph/plan-execute-and-verify-a-task/aiml.sh:build-a-multi-step-task-runner-with-independent-outcome-verificationBuild a LangGraph plan, execute and verify graph that applies a review fix with a create_agent executor and grades it by running the project's tests.Tue, 20 Oct 2026 00:00:00 GMTBuilding Agents with LangGraphhttps://aiml.sh/courses/mcp/build-a-remote-mcp-service-with-authentication-and-authorization/aiml.sh:build-a-remote-mcp-service-with-authentication-and-authorizationServe an MCP server over Streamable HTTP that accepts only tokens issued for it, publishes Protected Resource Metadata and returns 403 for a missing scope.Tue, 20 Oct 2026 00:00:00 GMTModel Context Protocol (MCP)https://aiml.sh/courses/mcp/build-a-scoped-postgresql-mcp-tool-with-read-only-roles-and-query-limits/aiml.sh:build-a-scoped-postgresql-mcp-tool-with-read-only-roles-and-query-limitsLet a model run SQL through an MCP tool on PostgreSQL 17 with a read-only role, a statement timeout and row and byte limits, and test what each layer stops.Tue, 20 Oct 2026 00:00:00 GMTModel Context Protocol (MCP)https://aiml.sh/courses/rag/build-a-vector-search-comparison-exact-ann-pgvector/aiml.sh:build-a-vector-search-comparison-exact-ann-pgvectorCompare pgvector exact, HNSW and IVFFlat results through PGVectorStore, check the query plan, and measure overlap on saved developer questions.Tue, 20 Oct 2026 00:00:00 GMTRetrieval-Augmented Generation (RAG)https://aiml.sh/courses/rag/build-a-vector-store/aiml.sh:build-a-vector-storeUse PGVectorStore to insert, replace, filter and delete internal documentation chunks, then test retries and a partially written batch.Tue, 20 Oct 2026 00:00:00 GMTRetrieval-Augmented Generation (RAG)https://aiml.sh/courses/llm-evaluation/build-an-eval-harness/aiml.sh:build-a-versioned-evaluation-harness-with-deterministic-checksBuild a versioned pytest eval harness for an LLM triage function with structured-output checks, a calibrated DeepEval G-Eval judge, pass^k and offline replay.Tue, 20 Oct 2026 00:00:00 GMTLLM Evalshttps://aiml.sh/courses/langgraph/checkpoint-and-resume-a-workflow/aiml.sh:build-a-workflow-with-checkpoints-and-resumable-stepsCheckpoint a LangGraph workflow to SQLite with SqliteSaver, kill the process mid-run, resume without calling Gemini again and see which steps repeat.Tue, 20 Oct 2026 00:00:00 GMTBuilding Agents with LangGraphhttps://aiml.sh/courses/rag/build-an-embedding-pipeline/aiml.sh:build-an-embedding-pipelineSplit internal developer documentation with LangChain, embed chunks through Ollama, and keep the model and chunk settings beside the vectors.Tue, 20 Oct 2026 00:00:00 GMTRetrieval-Augmented Generation (RAG)https://aiml.sh/courses/mcp/build-an-mcp-client/aiml.sh:build-an-mcp-client-with-discovery-and-execution-tracingBuild an MCP client that acts on discovery, refuses bad calls before sending them, classifies every outcome and traces each call with OpenTelemetry.Tue, 20 Oct 2026 00:00:00 GMTModel Context Protocol (MCP)https://aiml.sh/courses/mcp/build-an-mcp-server-with-tools-resources-and-prompts/aiml.sh:build-an-mcp-server-with-tools-resources-and-promptsBuild a docs server with the MCP Python SDK that offers a search tool, pages as resources and a review prompt, and returns the specified error codes.Tue, 20 Oct 2026 00:00:00 GMTModel Context Protocol (MCP)https://aiml.sh/courses/rag/build-exact-vector-search/aiml.sh:build-exact-vector-searchUse NumPy and saved real embeddings to inspect cosine distance, reject invalid vector shapes, and see why metadata filters still matter.Tue, 20 Oct 2026 00:00:00 GMTRetrieval-Augmented Generation (RAG)https://aiml.sh/courses/llm-evaluation/build-held-out-dataset-and-leakage-checks/aiml.sh:build-held-out-dataset-and-leakage-checksKeep related tracker issues in one split, review word overlap and copied prompt text, and detect changes after a held-out dataset is frozen.Tue, 20 Oct 2026 00:00:00 GMTLLM Evalshttps://aiml.sh/courses/langgraph/make-retries-idempotent/aiml.sh:build-idempotent-retries-for-workflowsRetry a ticket-filing write with a LangGraph RetryPolicy, see a retry file a duplicate, and stop duplicates across retries and resumes with an idempotency key.Tue, 20 Oct 2026 00:00:00 GMTBuilding Agents with LangGraphhttps://aiml.sh/courses/rag/build-multimodal-rag-with-verifiable-source-references/aiml.sh:build-multimodal-rag-with-verifiable-source-referencesRetrieve chart descriptions with PGVectorStore, send selected PNGs to Gemini, and check image identity separately from quoted source text.Tue, 20 Oct 2026 00:00:00 GMTRetrieval-Augmented Generation (RAG)https://aiml.sh/courses/llm-evaluation/gate-releases-on-eval-regressions/aiml.sh:build-regression-gates-for-prompts-models-data-and-indexesGate prompt, model and dataset changes by comparing a pytest eval run with the main baseline, block new failures and run it in GitHub Actions without a model.Tue, 20 Oct 2026 00:00:00 GMTLLM Evalshttps://aiml.sh/courses/llm-evaluation/measure-rag-and-tool-call-quality/aiml.sh:build-retrieval-generation-and-tool-execution-evaluation-reportsEvaluate a LangChain docs agent stage by stage with DeepEval, score its tool calls, retrieval and answer faithfulness, and report the first stage that fails.Tue, 20 Oct 2026 00:00:00 GMTLLM Evalshttps://aiml.sh/courses/llm-evaluation/build-robustness-tests-for-an-llm-app/aiml.sh:build-robustness-tests-for-an-llm-appExtend a pytest eval suite with meaning-preserving perturbations and prompt-injection cases, run it on Gemini and Gemma, and read per-perturbation pass counts.Tue, 20 Oct 2026 00:00:00 GMTLLM Evalshttps://aiml.sh/courses/langgraph/add-timeouts-and-token-budgets/aiml.sh:build-tool-timeouts-and-budgets-for-agentsBound a LangGraph agent with tool and model timeouts and a soft token budget, then inspect a genuine tool proposal followed by a native local timeout.Tue, 20 Oct 2026 00:00:00 GMTBuilding Agents with LangGraphhttps://aiml.sh/courses/ai-fundamentals/human-control/aiml.sh:human-controlTest whether document text can trigger an unauthorized tool call, and compare prompt warnings with application permission checks.Tue, 20 Oct 2026 00:00:00 GMTAI Fundamentalshttps://aiml.sh/courses/ai-fundamentals/tools-and-permissions/aiml.sh:tools-and-permissionsSend a read tool schema to a local model, inspect one proposed call, and reject unknown names or invalid page arguments before the tool function runs.Tue, 20 Oct 2026 00:00:00 GMTAI Fundamentalshttps://aiml.sh/courses/ai-fundamentals/vectors-and-similarity/aiml.sh:vectors-and-similarityEmbed the docuctl docs pages and questions with a local model, rank pages by cosine similarity, and see the two ways similarity search misleads.Tue, 20 Oct 2026 00:00:00 GMTAI Fundamentalshttps://aiml.sh/courses/ai-fundamentals/what-is-an-evaluation/aiml.sh:what-is-an-evaluationWrite fixed cases for the docuctl docs assistant, grade every answer in code, run three trials per case, and use the results to judge two attempted fixes.Tue, 20 Oct 2026 00:00:00 GMTAI Fundamentalshttps://aiml.sh/courses/ai-fundamentals/generation-controls/aiml.sh:generation-controlsSet temperature and seed, cap the output with num_predict, stream the reply and time its first chunk, and handle a reply that was cut off or timed out.Sun, 18 Oct 2026 00:00:00 GMTAI Fundamentalshttps://aiml.sh/courses/ai-fundamentals/json-and-files/aiml.sh:json-and-filesMake the docuctl docs assistant return a typed answer: send a Pydantic JSON schema as the output format, validate every reply, and see its limits.Sun, 18 Oct 2026 00:00:00 GMTAI Fundamentalshttps://aiml.sh/courses/ai-fundamentals/local-models-and-hosted-apis/aiml.sh:local-models-and-hosted-apisSend one docuctl docs question to Gemma through Ollama and to Gemini through its API, compare latency and token usage, and handle each failure.Sun, 18 Oct 2026 00:00:00 GMTAI Fundamentalshttps://aiml.sh/courses/ai-fundamentals/messages-and-context/aiml.sh:messages-and-contextBuild a multi-turn chat request for the docuctl docs assistant, resend the history Ollama needs, and detect when the configured context window loses input.Sun, 18 Oct 2026 00:00:00 GMTAI Fundamentalshttps://aiml.sh/courses/ai-fundamentals/next-token-prediction/aiml.sh:next-token-predictionRun a language model one token at a time, record the top candidates it scores at each step, and see how the docs in the context change every choice.Sun, 18 Oct 2026 00:00:00 GMTAI Fundamentalshttps://aiml.sh/courses/ai-fundamentals/tokens-and-tokenizers/aiml.sh:tokens-and-tokenizersTokenize docuctl docs, commands and IDs with Gemma's own tokenizer, then compare your counts with what Ollama actually processed, chat template included.Sun, 18 Oct 2026 00:00:00 GMTAI Fundamentalshttps://aiml.sh/courses/ai-fundamentals/what-is-inference/aiml.sh:what-is-inferenceRun a language model with and without supplied documentation, and explain how request context differs from trained model weights.Sun, 18 Oct 2026 00:00:00 GMTAI Fundamentalshttps://aiml.sh/courses/langgraph/trim-history-to-a-token-budget/aiml.sh:build-a-context-budget-managerTrim a LangGraph code-review thread with trim_messages, pin key decisions, and compare the token estimate with Gemini’s reported input usage.Sat, 17 Oct 2026 00:00:00 GMTBuilding Agents with LangGraph