AI
From Chat-with-PDF to Quiz-Master: Live-Grading RAG with LLM-as-Judge in Python
Coming soon: Moving beyond passive search, this live-coded session demonstrates how to build an interactive “exam engine” from complex documents. Learn how to combine layout-aware ingestion, synthetic QA generation, and an LLM-as-judge pipeline to move from basic retrieval to real-time, human-in-the-loop evaluation using Docling, DeepEval, and Marimo.
AI Signals #27: Not another LLM talk...Practical lessons from building a real-world adverse media pipeline
Presentation delivering a transparent characterisation of moving beyond early-stage proofs of concept to deploying large language models at scale. Drawing on experiences at Comply Advantage processing millions of multi-lingual articles daily, the talk highlights the nuances of parsing long-form text, managing model deprecation cycles, and establishing rigorous evaluation guardrails.
Bridging the Gap: Building Robust, Tool-Integrated LLM Applications with the Model Context Protocol
A hands-on technical workshop exploring the Model Context Protocol (MCP) client-server architecture, detailing how to build intelligent modular AI systems using standard Python toolchains.
The companies we keep: How robust data structures benefit us all
Presentation covering practical data science with the third sector on behalf of DataKind UK. Showcasing specifically the case study of the Companies We Keep project in partnership with Global Witness.
Technology and Data First Webinar
An executive panel discussion focusing on the strategic alignment, systemic barriers, psychological safety, and organisational transformation necessary when deploying new technologies like artificial intelligence for business and society.