Practical notes: How I Stopped Guessing and Started Measuring My RAG Pipeline

This walkthrough rebuilds the path from raw materials to a working system for: How I Stopped Guessing and Started Measuring My RAG Pipeline (LLM Zoomcamp 2026, Module 4). The focus is operable steps, explicit checks, and code that you can drop into a repo without guessing intent. For the Overview stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline.

Why Evaluation Is the Most Underrated Part of RAG

When working through the Why Evaluation Is the stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion. Cache stable system instructions and tool schemas. Re-sending identical preamble is a common source of burn.

Building a Ground Truth Dataset

When working through the Building a Ground Truth stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments. Cache stable system instructions and tool schemas. Re-sending identical preamble is a common source of burn.

The Two Metrics That Matter

When working through the The Two Metrics That stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph. Cache stable system instructions and tool schemas. Re-sending identical preamble is a common source of burn. When working through the The Two Metrics That stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline.

The Results That Surprised Me

The The Results That Surprised stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion. Budget tokens per turn and per session. Agentic tools expand context aggressively; hard caps keep demos from becoming surprise invoices.

The evaluate() Function That Changes Everything

The The evaluate Function That stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments. Budget tokens per turn and per session. Agentic tools expand context aggressively; hard caps keep demos from becoming surprise invoices.

def evaluate(search_func, ground_truth):
    hit_rates, mrrs = [], []
    for record in ground_truth:
        results = search_func(record["question"])
        relevance = compute_relevance(record, results)
        hit_rates.append(hit_rate(relevance))
        mrrs.append(mrr(relevance))
    return {"hit_rate": sum(hit_rates)/len(hit_rates),
            "mrr": sum(mrrs)/len(mrrs)}

the Full Solution

The the Full Solution stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph. Budget tokens per turn and per session. Agentic tools expand context aggressively; hard caps keep demos from becoming surprise invoices. The the Full Solution stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline.

Operational checklist

The Operational checklist stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope.

Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish.

Budget tokens per turn and per session. Agentic tools expand context aggressively; hard caps keep demos from becoming surprise invoices.

Cite the passages that actually grounded the answer. Without citations, operators cannot tell hallucination from an indexing gap.

Write a short runbook: how to rotate keys, how to drain the queue, how to roll back the last ingest.