1. Abstract
How do you design a data platform that can preserve years of history, support model development and stress testing, and still remain maintainable over time?

In this hands-on workshop, participants will work through the architecture and engineering challenges behind building a robust data warehouse for Risk and Finance use cases. Together, we will explore data modelling in BigQuery, ingestion of structured and unstructured data, AI-assisted extraction from PDFs, pipeline engineering and the design of a model development platform.

The session will combine discussion, problem-solving and practical work in GCP.

2. Agenda
Intro: Workshop objective, problem definition and requirements
Session 1: Architecture – constraints introduced by Big Query. Efficient data model design.
Session 2: Data ingestion – structured and unstructured data sources.
Session 3: Using AI to extract structured data from PDFs
Session 4: Pipeline Engineering
Session 5: Model Development Platform

3. Objectives
Design a platform that continuously ingests data from Risk and Finance systems, preserves history indefinitely, supports model development, stress testing and analytical replay, and remains maintainable for decades.

A practical session combining discussion and design work – building a robust data warehouse for model development purposes.

4. Target audience and Prerequisites
Software developers, data engineers and other technical professionals interested in data warehousing and platform design.  

5. Technical requirements
GCP account 
Personal laptop