Intermediate · 1 day
RAG in a Day
A focused workshop for developers: build a complete retrieval-augmented generation system in one day, from ingestion and chunking through citations and a first evaluation.
1 day · Intermediate · Workshops · Developers
Overview
Some teams need RAG in their hands before they need the full production course. This workshop is that day. Theory is kept short. The spine of the day is a working system: ingest documents, chunk them, embed, retrieve, prompt, cite and evaluate.
You implement each stage, see what a weak default does, and improve the pieces that most often fail. The result is not a complete production platform. It is a complete path a developer can extend, measure and take back to a product.
The workshop assumes professional developers. It does not spend the morning on what a token is. It spends it on a corpus, a question set and a pipeline that can be shown to a colleague.
By the end of the day, you have a RAG system you built, a small evaluation set, and a clear view of what would still be required for production.
Audience
Developers who need to implement a first serious RAG path and prefer to learn by building.
Prerequisites
Professional software development experience and the ability to run a small application locally. Prior LLM API experience is helpful.
Duration
1 day
Can be run on a customer corpus. Teams that need production architecture, ACLs and monitoring should follow with the two-day RAG course.
Learning outcomes
- 01
Implement document ingestion, parsing and chunking for a small corpus.
- 02
Generate embeddings and run vector search against a realistic question set.
- 03
Construct a grounded prompt with retrieved context and citations.
- 04
Evaluate a first question set and name the main retrieval failures.
- 05
Identify the production work still required after a working path exists.
Outline
- 01
Ingestion and chunking
- Corpus, document types and a question set for the day.
- Parsing and chunking that can be inspected.
- Build: ingest and chunk; look at the worst chunks.
- 02
Embeddings and retrieval
- Embeddings, an index, and similarity search.
- Filters and the first retrieval failures.
- Build: retrieve for the question set and score recall by eye and by simple metrics.
- 03
Generation, citations and evaluation
- Prompt construction, grounded answers and citations.
- Refusal when retrieval is empty.
- A first evaluation pass and a list of production gaps.
- Build: complete the path and demonstrate it on held-out questions.
Practical work
The entire day is a build. You leave with a working RAG pipeline covering ingestion, chunking, embeddings, retrieval, prompting, citations and a small evaluation set.
Takeaways
- A working RAG system
- A small golden question set
- Notes on the production gaps still open
- A path into the two-day RAG course if the team needs to go further
Delivery
Cognitecta delivers private corporate training, on-site or as remote live training. Courses can be run as published, or adapted to your organisation’s stack, domain and experience level.
Instructor
Nicholas Johnson, AI architect and software engineer. He has a degree in Artificial Intelligence and around twenty years of professional technology training, including hundreds of courses for engineering teams and large organisations. About.
Related courses
1 day · Intermediate
Build an AI Agent in a Day
A hands-on workshop for developers: build a working agent with tool calling, state, structured outputs, error handling, permissions and a first evaluation.
1 day · Intermediate
From Chatbot to Agent
A one-day workshop that starts with a basic LLM chat application and evolves it, step by step, into a useful agentic system with retrieval, tools, state and production safeguards.
1 day · Leadership
AI Architecture Workshop
A facilitated workshop for technical teams, architects and engineering leaders around a real customer problem. The output is an architecture and a next-step roadmap, not a training certificate.