The AI Forge · Foundation · F2
Retrieval & Knowledge (RAG)
Ingestion, chunking, embeddings, vector search, retrieval quality, deployed
- Duration
- 3 days · 9:00–17:00 each day (19.5h contact, incl. breaks + lunch)
- Participant level
- Intermediate: regular AI users
- Format
- Instructor-led labs, build-along
- Participants
- 8 included · 12 maximum (flexible group size)
- Prerequisite
- F1 (or equivalent)
- Ecosystem
- Agnostic: any model provider or stack
- Price
- EUR 11,500 (EUR 650 per extra participant)
Amounts in EUR, excl. VAT.
ai-generatedWhat participants will be able to do
- Design and build a production-grade RAG pipeline from scratch
- Choose and configure the right vector database for your use case
- Evaluate retrieval quality using precision, recall, and RAGAS metrics
- Deploy and monitor a RAG system in production
Tools & resources
What changes after this module
Engineers ship a complete, evaluated retrieval system that grounds a model in real data, and understand the retrieval patterns that persist as embedding models and vector stores change.
Who should attend
Engineers and data professionals. Intermediate: regular AI users.
Programme
This agenda is indicative. Content, sequencing, and examples are adapted to your team's context, tools, and objectives.
Before you start
- A laptop with a working development environment and admin rights
- Proficiency in at least one general-purpose language (Python and/or JavaScript/TypeScript)
- A code editor/IDE with an approved AI coding assistant enabled
- An API key for an approved model provider (issued by the company)
- A Git repository and command-line comfort
- Completion of F1 or equivalent
- A real document corpus to build against (anonymised if sensitive)
| Day 1 | 9:00–10:30 | When to ground, and how: Retrieval vs fine-tuning vs long-context; the RAG pipeline as a set of durable stages, not a specific framework Lab: Decide the right grounding approach for three real use cases and justify each on cost, freshness and accuracyOne becomes your build case for the three days. |
| 10:30–10:45 · ☕ Coffee break | ||
| Day 1 | 10:45–12:30 | Ingestion & parsing: Multi-format parsing, cleaning and metadata: the quality of what goes in caps everything downstream Lab: Build an ingestion pipeline that parses a mixed-format corpus into clean, metadata-tagged chunks. |
| 12:30–13:30 · 🍽 Lunch break | ||
| Day 1 | 13:30–15:30 | Chunking strategies: Fixed, semantic, recursive and structure-aware chunking; overlap; how to choose per content type Lab: Run the same corpus through several chunking strategies and score retrieval quality on each. |
| 15:30–15:45 · ☕ Coffee break | ||
| Day 1 | 15:45–17:00 | Lab: ingestion pipeline: Consolidate parsing + chunking into a repeatable pipeline Lab: Ship an ingestion pipeline, load your corpus ready for embedding, and draft the 20-question gold set you’ll evaluate against for the rest of the build. |
| Day 2 | 9:00–10:30 | Embeddings: What embeddings are, dimension/latency/cost trade-offs, multilingual concerns, benchmarking rather than brand-picking Lab: Benchmark two embedding approaches on your data and pick one on evidence. |
| 10:30–10:45 · ☕ Coffee break | ||
| Day 2 | 10:45–12:30 | Vector search: Indexing, similarity, metadata filtering and hybrid (keyword + vector) search: concepts stable across databases Lab: Stand up a vector store, index your chunks, and run hybrid queries with metadata filters. |
| 12:30–13:30 · 🍽 Lunch break | ||
| Day 2 | 13:30–15:30 | Retrieval optimisation: Re-ranking, query rewriting, hypothetical-document embeddings, and fusion: where most accuracy is won Lab: Add re-ranking and query rewriting and measure the accuracy lift over the naive baseline. |
| 15:30–15:45 · ☕ Coffee break | ||
| Day 2 | 15:45–17:00 | Lab: retrieval service: Wire embeddings + store + retrieval into a service Lab: Ship a retrieval service and measure precision/recall on a labelled query set. |
| Day 3 | 9:00–10:30 | Generation & grounding: Prompt design for RAG, citation/attribution, and reducing ungrounded answers Lab: Add grounded generation with citations and cut ungrounded responses on your eval set. |
| 10:30–10:45 · ☕ Coffee break | ||
| Day 3 | 10:45–12:30 | RAG evaluation: Faithfulness, context precision/recall, answer relevance: the durable RAG metrics; building the eval set Lab: Build a RAG eval suite and get a baseline score across faithfulness and relevance. |
| 12:30–13:30 · 🍽 Lunch break | ||
| Day 3 | 13:30–15:30 | Lab: complete RAG system: End-to-end ingest → embed → retrieve → generate, deployed Lab: Deploy the full RAG system and query it live against your corpus. |
| 15:30–15:45 · ☕ Coffee break | ||
| Day 3 | 15:45–17:00 | Production & review: Caching, rate limiting, monitoring, cost at scale; peer review Lab: Add caching + monitoring, then present your system with its benchmark numbers. |
Deliverables
- Deployed, queryable RAG system
- Ingestion + chunking pipeline
- Retrieval service with hybrid search + re-ranking
- RAG eval suite with baseline scores
- Production notes (caching, monitoring, cost)
Interested in running this module for your team? Get in touch and we'll tailor the format, dates, and delivery to your context.
Request this trainingYour trainer
Senior expert
Has shipped production RAG systems; deep retrieval, chunking and evaluation experience
Discovery Session
This module requires a full day to deliver a meaningful learning experience. We only offer it in its complete format.
Frequent questions
Can modules be taken individually?
Yes. Every module stands alone at a fixed price, and every module counts toward a programme if you continue.
Where does training happen?
At your premises or remote, on your dates, for private cohorts. Open sessions run on a fixed monthly calendar.
Which AI tools do you train on?
Yours. Every module ships in four ecosystem editions and runs its exercises on your real stack.
Who delivers?
Inforca's senior consultants and trainers. Flagged modules and the executive track are delivered at senior-expert level.
Discuss this module in a First Call: fit, dates, and the path around it.
No commitment · our team responds within one business day