Toby Drane (@toby_drane) header
Toby Drane (@toby_drane) profile picture

building AI @meetgranola previously AI @10downingstreet

London
Joined June 2014
Photo by Toby Drane (@toby_drane)Photo by Toby Drane (@toby_drane)Photo by Toby Drane (@toby_drane)Photo by Toby Drane (@toby_drane)Photo by Toby Drane (@toby_drane)Photo by Toby Drane (@toby_drane)
Thank you @meetgranola for having us today! Loved meeting everyone and seeing what cool and wild things everyone has built with Grok @bot 💚
11
6
2
179
7,953
Toby Drane retweeted
Dimension is joining Granola. After nearly four years of building, our mission found a bigger home. More details in the thread.
53
6
13
302
74,329
Toby Drane retweeted
On the day America celebrated the IPO of its biggest space company, Britain found out its government is banning underfloor heating. The decline of our country is not inevitable, it is the choice of the rotten political class.
386
2,534
71
18,355
478,387
Toby Drane retweeted
One of our meeting rooms is a dome on the roof. Perfect for a day like today ☀️ Oh, and also we're hiring.
8
5
3
73
13,719
Toby Drane retweeted
the more you Granola, the more you get.
9
9
12
225
142,244
Toby Drane retweeted
So @meetgranola released their official MCP server roughly a year after MCP became a thing. I just ran a side-by-side comparison against our my meeting intelligence pipeline — the results aren't even close. The design is exceptionally well thought out — particularly the semantic query tool with inline citations. This is a case study in "taking your time to do it right." The Past My old approach to ingesting meeting intelligence into Claude: 1. Query Notion database (Granola syncs there) 2. Get page IDs back 3. Fetch each page individually 4. Parse Notion markdown 5. Extract insights from raw text That's 4-5 API calls per meeting. For 19 meetings, you're looking at ~80+ calls. The Future With Granola's MCP: "What key decisions were made this week?" Returns: Synthesized intelligence with inline citation links to each source meeting. That's it. One call. Done. What makes this MCP exceptionally well-designed: • 4 tools total (list, query, get, transcript). No bloat. • Semantic search with citations built in — not bolted on • Participant data pre-resolved ("Laura from Descript", not raw emails) • Summaries are structured, not raw transcript dumps Granola didn't ship "get_note_by_id" and call it a day. They built semantic intelligence into the protocol layer. Bravo. Really impressed and already updated my agents (which was mostly a drastic reduction in instructions meant to manage the complexity of extracting info from the Notion db) and added a new skill so Claude Code can reference meeting discussions on demand. Here's what @claudeai had to say after testing both approaches on the same set of meetings:
13
15
6
304
64,407
Toby Drane retweeted
good morning san francisco
9
9
5
179
15,066
RT @meetgranola: Sign in with Microsoft is here! If you've been waiting to use Granola with your Microsoft account: now you can ✨ https://…
3
24
Toby Drane retweeted
tyvm @toby_drane for the idea
1
1
8
2,331
Toby Drane retweeted
We’re in the #BrexBenchmark Top 25 🎉 Brex’s aggregated, anonymised spend data shows Granola is one of the fastest-growing tools their customers are buying. Thanks for the recognition, @brexHQ 🤝
1
4
1
36
5,101
Toby Drane retweeted
Your 2025. Crunched. See yours now in the Granola app.
27
16
40
159
81,949
EU regulations will block Apple's new super helpful live translation feature on the new Airpods. The EU... the place in the world possibly most useful to have that feature. You starting to see why the economy's going to hell in a handbasket? politico.eu/article/apple-bl…
53
200
31
945
52,079
Atla Selene Mini is a SoTA small language model-as-a-judge (SLMJ) for general-purpose evaluation. It outperforms GPT-4o and other SLMJs across 11 benchmarks, excelling in scoring, classification, and preference tasks. Trained with DPO and SFT, it’s robust, promptable, and achieves high zero-shot agreement with human experts. - Top 8B model on RewardBench. - Available on HuggingFace & Ollama. - A larger version is coming 🔜
1
3
9
564
Toby Drane retweeted
Atla Selene Mini A General Purpose Evaluation Model
1
5
26
9,076
Congrats to the Atla team, so nice to see this work launched.
This Post is from an account that no longer exists. Learn more
2
9
1,295
Under the leadership of @Keir_Starmer @RachelReevesMP @MayorofLondon: The UK has developed a skill in imports and exports. We import criminals and export talented entrepreneurs.
97
137
20
1,539
94,314
To summarise the LFG event: - this is a movement, join our dinners to get involved - we have a bill ready that will fix our energy and compute problems, help us send physical copies to ministers on the website - the country is broken, but fixable, help us - we need actual talent in politics, join us
12
19
11
297
120,897
⚖️ Judge Arena: Benchmarking LLMs as judges! I'm currently having fun judging LLMs judges :), this is going to be quite useful too! huggingface.co/spaces/AtlaAI…
3
12
1
64
8,759