Growth @HAQQ Legal AI | Legal Intelligence for Everyone

Paris, France
Joined November 2023
Could the recent launch of Intake by @WeAreLegora and @harvey be a sign of a merger?
22
1/ I've never taken a course on AI. Here's everything I actually do, from using it every day.
1
34
11/ Keep a notes file. Anything I see on Instagram, X or LinkedIn goes in. Each morning I hand it to Claude and let it run.
1
9
12/ And one day you stop renting. I've built my own setup. I know what to expect from it. That only comes from using it daily. Start with tip 2.
2
Introducing Intake: the agentic legal front door for in-house teams. Every in-house legal team runs on requests from the business. A contract to look at, a supplier to approve, a policy question to answer. They arrive constantly, in threads and inboxes, often half-scoped, and the real legal work starts only after someone has read the request, chased the missing attachment and logged it somewhere. Intake changes that process. It turns ad-hoc requests into tracked work in Legora, routes the execution to the Legora Agent, and returns it to the requester, with a lawyer accountable the whole way through. Available now for in-house teams. Learn more: legora.com/product/intake
37
The cuter the machine, the further inside it gets. Fields, then factories, then your home, then your laptop and bank account. Natalie Fratto made a Threat to Pet Matrix for AI agents and robots. I turned it into an interactive map with the research behind it.
1
65
The scary part is overtrust. Georgia Tech, 2016: in a smoke-filled evacuation drill, all 26 participants followed a guide robot. Half had just watched it fail a simple navigation task. Trust should track performance. Looks move it anyway.
1
27
If you build agents: → Match the face to the proof → Stylise, don't simulate → Let the face show uncertainty → No guilt-trips when users leave → Plan the shutdown Place your own agent on the map 👉 claude.ai/artifact/53CxhDhow…
25
jev is amazing 😄
2
77