A comprehensive, advanced GenAI training for data science teams ready to move from experimentation toward production-relevant systems – extended with retrieval-augmented generation, reasoning patterns, and RAG pipeline evaluation on company data.
This three-day workshop expands the earlier foundations with retrieval-augmented generation, reasoning patterns, and more structured approaches to designing, implementing, and evaluating AI workflows on company data.
Generative AI field update
Understand what is new and relevant in the GenAI landscape - grounded in practical engineering, not hype. Focus on what actually matters for data science workflows in 2026.
Coding with AI agents
Hands-on introduction to building with AI Agents. Each participant works in a clean, isolated workspace to ensure safe and reproducible learning outcomes.
LLM integration patterns
Practical patterns for integrating Large Language Models into data science applications. Covers common mistakes, cost implications, and reliability considerations.
Evaluation & quality basics
Learn how to assess and measure LLM output quality. Build risk awareness and understand reliability constraints before committing to larger architectural investments.
DURATION
3 full days: Instruction + exercises
FORMAT
Online workshop: Clean workspace per participant
GROUP SIZE
6 – 12 Recommended participants
PREREQUISITES
Python & CLI + data science experience
Six core topics covered across the day, building from fundamentals to practical implementation.
This workshop is designed for data science teams who have moved beyond initial LLM experiments and are ready to design, implement, and evaluate production-relevant AI systems on company data.
Data scientists
Teams ready to go beyond agentic workflows and implement RAG systems using their own organisational data - with rigorous evaluation and reliability built in.
ML engineers
Engineers designing the architecture of production AI systems who want deep understanding of RAG patterns, retrieval evaluation, and reasoning strategies.
Technical leads
Team leads committing to a full AI capability build - wanting their team to leave with production-ready skills applicable to the organisation's real data and systems.
Participants will be able to apply agentic development to their own business challenges, implement RAG systems using company data, and incorporate reliability and evaluation into their solutions. They’ll leave with a broader architectural understanding of modern GenAI systems and the ability to connect RAG design, reasoning strategies, and evaluation thinking in production contexts.
The 1-day workshop is a structured entry point covering GenAI fundamentals. The 2-day format adds tailored exercises on client datasets and deeper agentic development. The 3-day format includes everything from both, then adds a full third day dedicated to Retrieval-Augmented Generation – designing and evaluating RAG pipelines on company data, and structured reasoning strategies. It is the most comprehensive option and is recommended for teams ready to build production-ready AI systems.
No prior TSW Academy course is required. Days 1 and 2 of the 3-day program cover all the foundational and applied content from the shorter formats. However, teams with prior hands-on GenAI experience may find the pace of days 1–2 faster, which leaves more room for depth on day 3. If your team already has solid GenAI foundations, we can discuss adjusting the balance during the discovery phase.
Yes, a discovery phase before training aligns the program with your team’s technical maturity, tooling, datasets, and goals. Days 2 and 3 exercises are built specifically around your organisation’s context and data. This makes the 3-day format the most deeply customised option in our offering.
Participants should be comfortable with Python, the command line interface, and have prior data science experience. No prior GenAI or LLM experience is required – day 1 covers all foundational concepts. For the RAG content on day 3, familiarity with basic data storage or document management concepts is helpful but not required.
TSW Academy’s programs combine engineering discipline with architectural thinking. We focus on realistic system behaviour, common failure modes, and production readiness — not just how to make impressive demos. Our instructors are practitioners, not generalist trainers.
You may cancel or reschedule free of charge up to three weeks before the scheduled date. Any preparation work already completed may be charged in the case of a late cancellation. The proposal is valid for 90 calendar days from issue.