Applications, Production, and Compound AI Systems

Covers creative writing, content generation, text transformation, generation quality control.

Covers extractive vs abstractive, length control, faithfulness, multi-document summarization.

Covers open-domain QA, reading comprehension, knowledge-intensive QA, QA evaluation.

Covers named entity recognition, relation extraction, event extraction, structured output generation.

Covers sentiment analysis, intent detection, topic classification, zero-shot classification.

Covers dialogue management, context tracking, persona consistency, conversation evaluation.

Covers story generation, poetry, code creativity, creative constraints and control.

Decomposing systems into components, explicit orchestration, typed interfaces, and failure boundaries.

Model cascades, confidence-based routing, deterministic fallbacks, abstention, and partial service.

Task suites, traces, regression gates, error taxonomies, and improving systems without benchmark theater.

Prompt, retrieval, tool, latency, cost, and quality traces with privacy-aware retention.

Coordinated releases, compatibility, reproducibility, canaries, rollback, and artifact lineage.

Service objectives, retries, idempotency, concurrency, capacity, caching, budgets, and failure testing.

Covers serving frameworks, model loading, request handling, serving configuration.

Covers latency breakdown, batching latency, streaming responses, latency monitoring.

Covers batch size tuning, GPU utilization, concurrent requests, throughput measurement.

Covers scaling metrics, horizontal scaling, scale-up vs scale-out, scaling policies.

Covers model selection, A/B testing, model cascades, routing strategies.

Covers prompt caching, semantic caching, cache invalidation, cache hit rates.

Covers metrics collection, alerting, logging, dashboards.

Covers output quality metrics, drift detection, regression detection, quality alerts.

Covers cost modeling, cost optimization, cost allocation, cost monitoring.