SAIC#032 – RAG in Practice: Navigating Advanced Techniques and Real-World Pitfalls
While basic Retrieval-Augmented Generation is a great start, scaling it for business requires more than just a vector database. This session dives into advanced architectural patterns, such as hybrid search, query transformations, and reranking, designed to solve the "hallucination hurdle" in professional settings. We will explore which techniques best fit specific use cases, how to navigate common implementation pitfalls, and what to do when standard approaches fall short in real-world situations.
SAIC#033 – Secure AI Agent Use Cases in Enterprise Environments
AI agents are moving from demos to day-to-day workflows, but deploying them inside a regulated enterprise raises harder questions: how do you capture the productivity gains without breaking data governance, identity management, or auditability? Mira El Kamali and Elmira Gazizova share their journey deploying AI agents in enterprise environments and the hybrid approach they adopted to balance automation, security, and human oversight, with concrete lessons on integrating agents into existing workflows and keeping humans in the loop where it matters.
Sold out. Three tracks, one evening — pick the path that fits your level. Claude Cowork (no coding required): automate file and task management from your desktop. Claude Code — Getting Started: your first steps with Claude in the terminal, from setup to building something real. Claude Code — Advanced Concepts: CLAUDE.md configuration, multi-file workflows, worktrees, skills, sub-agents, MCP servers, and production patterns. Bring your laptop (Mac or PC) with Claude Code already installed if you're joining a Code track. Spots limited to 20. Dinner and drinks on us. Hosted by Exponent.
Past Events
SAIC#031 – From Scarcity to Sovereignty: What is Civil Society's Role in Driving AI Development?
As AI accelerates, the organisations that represent citizens and the common good are being left behind, underfunded, under-resourced, and largely absent from influencing and making decisions. Why and how is this happening, and what can we do to change it?
SAIC#030 – Harness Engineering: What We've Learned Letting AI Agents Write Our Code
When AI agents write most of your code, you'd think the job gets easier. It does, and it doesn't. You spend less time implementing and more time setting up your repo, docs, and review gates so the agents don't produce garbage. OpenAI recently called this "harness engineering". Nico (co-founder of Velario) and Denis (AI Engineer) will talk about what this looks like day-to-day with Claude Code: how small teams actually ship this way, which workflows hold up, and what falls apart.
SAIC#029 – Your Company Is a Model. What Happens When AI Agents Start Optimising It?
François Chollet, a researcher at Google, noticed agentic coding looks like ML: define objective, set constraints, iterate, get a black box. Elliot Vaucher (co-founder at Ogram) applies that lens to law, finance, marketing, ops — and shows what breaks when your org becomes the thing being optimised.
SAIC#028 – No Code, No Problem, All is Possible — My AI & Lovable Journey
How Antoine Jasnault really discovered AI, what AI can really help achieve in an industrial environment with limited budget and high regulation constraints — without any coding knowledge. A very personal journey from discovering to building things with the Lovable platform.
SAIC#027 – GenAI in Banking: Accuracy, Compliance and Usability
We explored how banks deploy GenAI while navigating compliance, accuracy, and usability trade-offs — from on-premise small language models for CV screening to SQL-based guardrails against hallucination.
SAIC#026 – The Art & Science of Building Large-Scale Agentic Systems
How a Head of AI built the cruise industry's first large-scale GenAI customer service platform — 99.5% accuracy, tens of thousands of conversations weekly.
Past Events
SAIC#025 – Why Multi-Agent? Building AI Teams That Actually Work
Explore the power of multi-agent AI systems and learn how to build collaborative AI teams that deliver real results. This event covers practical approaches to orchestrating multiple AI agents for complex workflows.
SAIC#024 – BMAD Method: Turn AI into a Product Team
Discover the BMAD (Build, Measure, Analyze, Deploy) framework for context engineering. Learn how to structure AI systems like a product team and create more effective AI applications.
Deep dive into AI-powered financial analysis and trading systems. Learn how practitioners are using AI to analyze markets, identify opportunities, and make data-driven investment decisions.
