KNOWLEDGE LAB / LEARN RAG THROUGH PRACTICE
Build answers from your documents
Start with one document and one question. Learn to retrieve evidence and check the answer.
New to RAG? Start with one file
Use the sample FAQ to find a booking rule, cite its source and handle a question the document cannot answer.
Start with a one-file example Explore runnable concept labsChoose your RAG use case
The goal shapes your learning path. Your document format comes next.
THE PRACTICE LIBRARY
RAG tutorials: from retrieval to evaluation
Read worked examples with source files, expected outcomes, and troubleshooting checks. Browse independently or follow your learning plan.
Start with these three lessons
Build your first answer from one file
Prepare a practice file, run local retrieval, and connect the returned evidence to your chosen answer model.
Add and verify source citations
Carry source IDs through search and generation, then validate each citation against the passage.
Handle questions without evidence
Add a no-evidence path and test it beside a covered question.
Continue with lessons 4–6
RAG Debugging Starts by Tracing Where the Required Fact Disappears
Trace one question from source text to index, search result, model input, and final answer.
RAG Top K Should Be Chosen by Evidence Coverage and Final Answer Quality
Compare one returned passage with three while keeping documents and questions fixed.
Current RAG Answers Need New Evidence Indexed and Old Versions Retired
Replace a 60-minute rule with 90 minutes and remove the old version from searchable results.
GO DEEPER WHERE YOU NEED TO
RAG configuration: documents, models and databases
Explore document preparation, model and database choices, agent behavior, and fair comparisons.
Browse 8 configuration guides
RAG Retrieval and Answer Generation Solve Different Problems
Use the printer and parking exercises to separate retrieval failures from answer-model failures, then record a reproducible model choice.
The Right RAG Database Depends on Your Stack, Updates and Access Rules
Compare a local baseline, PostgreSQL with pgvector, and a dedicated vector service using source records, version updates, and access-filter exercises.
A Fixed RAG Workflow Can Be Enough Without an Autonomous Agent
Separate the answer model from its workflow, compare direct orchestration with LangGraph, and study a bounded multi-search example.
A Working RAG Knowledge Base Needs Checks from Import to Final Answer
Go from a small document set to a cited answer, with a check at every stage.
Product FAQ Retrieval Needs Identifiers and Exceptions Beside Each Answer
Keep product identifiers, exceptions, and versions attached to the answer.
Long-Manual RAG Must Preserve the Conditions That Change an Answer
Compare section boundaries and parent-child retrieval without losing conditions or exceptions.
OCR Errors Can Break Scanned-PDF RAG Before Retrieval Even Starts
Verify OCR, reading order, and table structure before spending time on retrieval settings.
RAG Evaluation: Strong Retrieval Scores Do Not Prove Correct Answers
Use paired questions, source checks, and measured usage to compare two configurations.
References & optional worksheets
Use these when a lesson calls for a closer look at configuration, a failure, or a measured comparison.