A working day for analysts, quant researchers, and PMs who want to scale their research with AI, not just experiment. The gap is rarely the model - it's the context you feed it and the control you keep over it. You'll brief AI on a real name and pressure-test its draft against the filings, point it at a whole coverage universe, turn unstructured sources into an event study, and ship a governed tool live. Hands-on throughout, led by AI-trained practitioners. 

Across 4 hours, you will explore 4 distinct modules:

Module 1 - First Draft, Faster: Single-Name Research from Filings to Note

Building an agent that drafts first-pass notes, earnings previews, and meeting-prep packs. We'll assemble a real name's source set, write the brief, generate a draft, and pressure-test it against the filings to catch what's plausible but wrong. Leave with a reusable briefing template and a verification checklist.

Module 2: Coverage at Scale: An Agent That Watches Your Universe 

Building an agent that watches a coverage list and surfaces what needs attention. We'll set the universe and inputs, define what counts as material versus noise, design the daily briefing and alerts, then add escalation rules for when the triage is trusted versus when a human steps in. Leave with a working monitoring agent and a triage rubric.

Refreshment & Networking Break

Module 3 - Reading the Reaction: Event Studies from Unstructured Sources

Building a workflow that turns unstructured sources into a structured event dataset and measures price reactions. We'll extract and validate events, join them to a price feed, and interpret the result

Module 4 - From Prototype to Piloted Tool: Shipping and Containing an AI Agent

Taking one of the afternoon's prototypes and deploying it as a real tool on Railway, then adding the governance layer - guardrails, monitoring, and a trigger for pulling the tool if it drifts.

Evening Networking & Drinks Reception