Course · Building RAG Knowledge Assistants

RAG Training: Build AI Assistants That Answer from Company Documents

An assistant that answers from your organisation's own documents is the production use case people ask about most. This course is for technical teams who will build and run that system, and covers answer quality, access permissions and evaluation.

Level
Advanced
Length
1 day
Hours
09:00–16:00
Tools
Model APIs and a vector database
Format
Onsite or online, hands-on

This course runs in-house at your organisation, taught in English or Thai, with cases and exercises rebuilt from the work of the department taking it.

Request a quoteAll courses

What you will be able to do

  • Design a pipeline from raw documents to answers with citations
  • Prepare and chunk documents so retrieval is accurate
  • Set up retrieval and ranking, and know what to adjust when an answer is wrong
  • Enforce access permissions so users see only the documents they are allowed to see
  • Evaluate answer quality systematically, before and after going live

One-day schedule

09:00–16:00 · topics, cases and exercises can be adapted to your organisation’s work · can be extended to two days for a longer workshop on real work

  1. 09:00–10:30

    RAG and preparing documents

    • What RAG is, and the problems it solves that fine-tuning does not
    • Chunking documents, and metadata
    • Updating documents when a new version arrives

    ActivityDesign a pipeline from raw documents to answers with citations

  2. 10:30–10:45

    Break

  3. 10:45–12:00

    Retrieval and generating answers

    • Embeddings, ranking and hybrid search
    • Generating answers with citations, and stopping the model from making things up
    • What to adjust when an answer is wrong

    ActivitySet up retrieval and ranking on a sample document set

  4. 12:00–13:00

    Lunch

  5. 13:00–14:30

    Access permissions and evaluating quality

    • Enforcing permissions so users see only the documents they are allowed to see
    • Sensitive data in documents
    • Evaluating answer quality with a test set

    ActivityBuild a question-and-answer test set, and measure quality before and after tuning

  6. 14:30–14:45

    Break

  7. 14:45–16:00

    Going live

    • Cost, latency and running the system
    • Monitoring quality after going live

    ActivityWorkshop: an AI assistant built on your organisation's real documents

Before the class

  • A team that can write code and understands basic APIs

Frequently asked questions

What background do we need?
The team must be able to write code and understand basic APIs. The course sits at level 4 of the AI Learning Layer and is for technical teams only.

Contact

Talk to us about training your team

Send a rough brief — how many people, which functions, and what they should be able to do afterwards — and we reply with a format and a quotation. We correspond in English or Thai; quotations and documents are issued by Trainiac Co., Ltd. for your procurement process.

Email
hello@trainiac.ai
Phone
061-894-0404
LINE Official Account
@trainiac
Email us