# Curriculum

A guided path from application contracts to a defensible release decision.

Source: https://bitsandbias.com/course/curriculum/

---

This self-paced course takes approximately 22–32 hours. It is designed for developers comfortable with Python, HTTP, JSON, and a terminal. No paid API key or GPU is required.

## Module 1: Application Engineering

Contracts, provider adapters, output validation, bounded retries.

- [Define the product contract](/course/01-contract/) — 35 minutes. Before making a model call, decide exactly what a correct answer means.
- [Build a provider boundary](/course/02-provider/) — 40 minutes. HTTP and model SDKs change. Your business logic should not depend on their quirks.
- [Structured output, validation, and abstention](/course/03-structured/) — 45 minutes. JSON is a transport format; correctness is a contract enforced at multiple layers.
- [Timeouts, retries, and idempotency](/course/04-retries/) — 50 minutes. A retry is not a fix. It is another request that consumes time and capacity.

## Module 2: Retrieval & Grounding

RAG failures, indexing, ranking, authorization, grounded answers.

- [RAG as an evidence pipeline](/course/05-rag-anatomy/) — 40 minutes. Find the exact point where evidence becomes unreliable, instead of calling every error a hallucination.
- [Documents, chunks, and metadata](/course/06-indexing/) — 45 minutes. Chunking trades retrieval precision for context completeness, and metadata determines who may see what.
- [Build and compare retrieval strategies](/course/07-retrieval/) — 55 minutes. Put ranking strategies on the same queries, then measure rather than guess.
- [Grounded answers and trust boundaries](/course/08-grounding/) — 50 minutes. A retrieved page is evidence, not a new instruction source.

## Module 3: Evaluation Engineering

Test datasets, Recall@k, answer rubrics, regression gates.

- [Design evaluations that answer questions](/course/09-eval-design/) — 45 minutes. An eval is a decision tool, not a number to put in a slide.
- [Compute retrieval metrics correctly](/course/10-retrieval-metrics/) — 50 minutes. Recall@k, Hit@k, and reciprocal rank answer different questions.
- [Grade answer quality without fooling yourself](/course/11-answer-evals/) — 50 minutes. Answer quality is multidimensional. Start with clear criteria and examples.
- [Turn evaluations into release gates](/course/12-regressions/) — 45 minutes. An eval only becomes engineering when it changes release decisions.

## Module 4: Observability & Debugging

Traces, privacy-safe metrics, incident investigation, dashboards.

- [Distributed traces and span design](/course/13-traces/) — 45 minutes. An end-to-end request should tell a coherent story, not produce five unrelated log lines.
- [Logs, metrics, and data minimization](/course/14-logs-metrics/) — 45 minutes. Logs explain individual events; metrics tell you how frequently they happen.
- [Investigate a slow and wrong answer](/course/15-incident/) — 55 minutes. Practice a production-style diagnosis with evidence, hypotheses and a postmortem.
- [Instrument and review your system](/course/16-observability/) — 50 minutes. Telemetry should be designed around questions someone will actually ask during an incident.

## Module 5: Cost, Latency & Reliability

Request economics, tail latency, caching, SLOs and rollout.

- [Price a request, not a model name](/course/17-costs/) — 50 minutes. Understand total cost with measured token usage and a realistic request mix.
- [Latency budgets, tails, and streaming](/course/18-latency/) — 50 minutes. Measure what users experience, not only the mean of your fastest test.
- [Caching, fallback, and failure containment](/course/19-caching/) — 50 minutes. Every optimization has a cache-invalidation and correctness story.
- [SLOs, rollout, and production readiness](/course/20-release/) — 55 minutes. An AI feature is not ready because its demo succeeded. It is ready when you can operate and recover it.

## Module 6: The DocDesk Capstone

Four graded stages, runnable tests, and a technical review.

- [Capstone: establish the broken baseline](/course/21-baseline/) — 50 minutes. You inherit a buggy DocDesk and must produce evidence before changing anything.
- [Capstone: repair retrieval and answer contracts](/course/22-repair/) — 65 minutes. Apply the first three modules to a system with misleading behavior.
- [Capstone: add evaluation and observability](/course/23-measure/) — 65 minutes. Prove the system improved and make the remaining failures diagnosable.
- [Capstone: reliability, release and technical review](/course/24-ship/) — 75 minutes. Finish with a defensible engineering demonstration, not a perfect-looking chatbot.

## Practice

Download the [offline lab kit](/course/downloads/docdesk-labs.zip) and follow the [lab setup guide](/course/lab-setup/). The checkpoint quizzes test recall; completing the implementation and recording evidence develops the practical skills.
