Today we just launched our Insights product Intercom My favourite feature: when a human answers a question that Fin couldn't, we use LLMs to suggest a change to the original content. Low friction, maintainable, powerful - the start of a genuine CS AI flywheel.

3 - Fin Suggestions (my fave) I ❤️ this workflow. Fin will: 1. Observe what happens when it escalates 2. Detect what's missing/wrong in your help centre 3. Suggest a diff on the article or a new article 4. You accept+publish, one click 5. Fin will never need to escalate again

Aside from our product updates we’ve been pretty quiet about what’s going on at Intercom and Fin as a business. Until now. Eoghan McCabe shares some exciting highlights in this post, have a read👇

In this podcast with Mark from AI Ireland, I talk about Fin, the AI agent we've built, and how we run the AI Group at Intercom. In this snippet I talk about the specific qualities we look for in our AI scientists and engineers, and how we intersect AI and product development.

We're putting our money ($1M) where our mouths (or more specifically our resolution rates) are with Fin - read Eoghan McCabe's post for the details.

Today, we at OpenAI achieved a milestone that many considered years away: gold medal-level performance on the 2025 IMO with a general reasoning LLM—under the same time limits as humans, without tools. As remarkable as that sounds, it’s even more significant than the headline 🧵

x.com/i/article/1947…

We've a significant Fin announcement in 20 minutes! Watch it here ⬇️ x.com/i/broadcasts/1…

Fin is doing 1M resolutions a week - an exciting milestone. These are times when it successfully resolves an end-user support query. Fin's a real AI agent, delivering real value, in a way that helps businesses, helps their customers. Next up, 1M a day!

Every winning company feels like a clown car inside. The important thing is to keep it pointed towards a goldmine.