1. Home
  2. Workshops
  3. Building a RAG Chatbot on Your Data

Building a RAG Chatbot on Your Data

Over two days you build a chatbot that answers questions from your own documents using retrieval-augmented generation. You cover document ingestion, chunking, embeddings, retrieval, answer generation with source references and systematic evaluation. We also address access control and data protection, since a chatbot on company data carries real responsibility.

Request without obligation

Deutsche Version

Duration
2 days
Level
Intermediate
Format
Remote or on-site, English or German
Group size
Up to 12 participants per trainer

Who it is for

  • Developers building knowledge assistants and support bots
  • Data and ML engineers moving prototypes toward production
  • Tech leads evaluating internal chatbot projects
  • Architects who must align such projects with security requirements

What your team will be able to do

  • Explain how retrieval-augmented generation works and where it fails
  • Ingest and chunk documents from typical sources
  • Set up retrieval with embeddings and improve it with metadata and re-ranking ideas
  • Generate answers that cite their sources and decline when evidence is missing
  • Build an evaluation set and measure retrieval and answer quality
  • Respect permissions so users only see what they are allowed to see

Agenda

  1. RAG fundamentalsUnderstand the pipeline from documents to answers and the typical failure modes.
  2. Ingestion and chunkingLoad your documents, clean them and try different chunking strategies.
  3. RetrievalBuild a vector search, add metadata filters and examine retrieval results critically.
  4. Answer generationCompose prompts with context, return source references and handle questions without good evidence.
  5. EvaluationCreate test questions and measure the effect of every change on quality.
  6. Access control and privacyApply document-level permissions, discuss data protection and plan logging responsibly.
  7. DeploymentWrap the system in an API and a simple interface and define monitoring and update routines.

Prerequisites

  • Solid programming experience, preferably Python or TypeScript
  • Basic familiarity with REST APIs
  • A set of non-confidential sample documents or a sandbox corpus

Why apprat.io workshops

  • Taught by practising AI Code Engineers
  • Your repositories or a sandbox project
  • AI speed, human responsibility: review, security and governance included
  • ISO 9001 & ISO/IEC 27001 certified processes
  • Follow-up office hours after 30 days

Request without obligation

Tell us about your team and goals. We reply within one business day with a tailored agenda and a quote – free and without obligation.

24/7 call support in English & German:
+49 761 519 751 50

Frequently asked questions

Do we need our own documents?

Sample documents you may share are ideal. If none exist, we prepare a neutral sandbox corpus.

Which vector database do we use?

We choose a common option that fits your stack. The concepts transfer to other stores.

Is the result ready for production?

You get a working, evaluated reference implementation. Production use additionally needs your security, data protection and operations review.

Fields marked * are required.

Related workshops

All workshops