Fergal Reid
@fergal_reid
Followers
2K
Following
937
Media
50
Statuses
524
Chief AI Officer @intercom, building Fin. ML PhD, degrees in CS, Math. Interests in AI, product development, science, math, progress, flourishing.
Dublin City, Ireland
Joined September 2009
December 3rd, we're bringing our Generative event series to Ireland! 🇮🇪 If you're a founder, designer, engineer, researcher working across AI, join us in conversation with @fergal_reid who runs AI at @intercom and our very own Eoin O'Mahony who runs AI... at @lightspeedvp :)
2
1
11
Something I’ve been thinking about as we build Fin at Intercom 🧵 AI product development is weirdly tricky. You can keep the same UI, ship “no visible changes,” and still make massive progress under the hood. The challenge is knowing when to keep polishing vs. when to build
2
3
27
Proudly announcing our entry in the "can't figure out caching" world championships.
@Fin_ai (Intercom’s AI Customer Agent) has officially crossed 10 trillion tokens processed in production. This excludes backtesting, development traffic, and custom model usage. In @intercom, we split workloads across multiple vendors to optimize for cost, performance and
1
0
8
.@eoghan, CEO of @intercom, and @fergal_reid, Chief AI Officer, join @labenz on @cogrev_podcast to discuss how they built Fin—the AI customer service agent that went from 35% to 65% resolution rate over 30 months while pioneering outcome-based pricing at 99¢ per resolution. They
3
7
19
I always enjoy @destraynor’s no-nonsense insights. He and I discuss growing @Fin_ai to $50 million revenue run rate, his “four horseman” rubric for investing in AI companies, advice for product marketers, and why you shouldn’t split the G. Subscribe on Spotify, Apple Podcasts,
23
25
288
Interested in a candid deep dive on what's actually required to get AI systems to take consequential actions? I sat down with @pratik60 to discuss his work lead our Tasks AI workstream building the Fin Tasks product. Link below!
2
1
6
Myself and Fedor Parfenov recently recorded a podcast about his work leading the Fin Insights workstream. Lots of interesting details about our journey to build a next generation LLM powered analytics product, and measuring success using Causal Inference. Link below!
1
2
9
Last night in San Francisco we brought together AI experts from Fin, @cognition, @harvey, and @perplexity_ai to share their experience building frontier AI products. It was a full house and the energy in the room was palpable – people building, with optimism and ambition. Thank
1
4
20
Myself and @thoughtwax made another podcast about the overall state of play of AI today, and AI strategy. If anyone is looking for incredibly niche technical content, check out the link in thread!
1
2
15
Really excited that Brett Chen from @perplexity_ai will be joining us as a speaker at our AI event next week. He'll talking about about "Scaling Intelligence: Production-Grade AI Agents and Models at Perplexity". Live in SF, and streaming. Details in thread.
1
5
20
Really excited about this event we're doing in SF with @cognition and @harvey, about building leading AI products. @silasalberti and @nikogrupen have kindly agreed to speak about their experiences, and I'll also be talking about our journey with @Fin_ai
https://t.co/3OUnPJFEXr
0
9
34
Every winning company feels like a clown car inside. The important thing is to keep it pointed towards a goldmine.
0
0
7
Our newest entry on the Fin research blog is by Afonso Marques. It's a deep essay arguing that LLMs already have sufficient peak intelligence to automate real world tasks like customer service, and that intelligence _density_ is what's required next.
1
0
13
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!
Fin just passed 1M resolutions / week (Up from 600k a quarter ago) 1 resolution = 1 customer query that’s been successfully resolved in the eyes of the customer. Which often requires back and forth and work across multiple systems Using the following numbers from our customer
3
1
27