Building a company is chaotic.
Your finances don’t need to be.
T:0 is an always-on agentic finance team for all the "quick questions" you’d never bother your accountant with.
Now generally available.
Breaking news: a company launched an MCP server. 🙃
We know. But ours lets you ask your AI accountant "what's my cash position?" or "who do I need to pay this week?" and get an answer without opening another dashboard.
To prove it isn't boring, we had Claude turn our financial statements into a retro arcade game.
Breaking news: a company launched an MCP server. 🙃
We know. But ours lets you ask your AI accountant "what's my cash position?" or "who do I need to pay this week?" and get an answer without opening another dashboard.
To prove it isn't boring, we had Claude turn our financial statements into a retro arcade game.
"I wish when I started @airwallex 10 years ago, T:0 existed.”
Even elite fintech founders hate doing their own accounting. That's exactly why we built T:0.
@awxjack@lancectk
A lot of accounting platforms claim to have "native integrations."
Behind the curtain? They are just manually downloading and uploading your CSV files on the backend and calling it automation.
Not at T:0. We build true, API-level integrations with the modern fintech stack you actually use. Your data flows instantly without the operational debt or fake automation.
Keenable Search API takes #1 spot on Vals Web Search Index
On GPT-5.6 Sol, Keenable pushes the quality frontier from Exa’s 45.24% to 49.08%, at less than half the cost per task
Get started with Keenable, first 100K requests on us
We’re partnering with @OpenTable 🍽️
OpenTable connects diners around the world with 70,000+ restaurants, many of them neighborhood businesses. We’re excited to help them thrive across markets.
Starting in the US, and Canada, with Australia and Mexico planned later this year.
As many of you know, over the past two years I’ve shifted my research from foundation models and NLP into AI-native chip design. What I’ve found: the problems feel 10x harder than software engineering. Writing functionally correct RTL is only the beginning.
As models saturate on benchmarks like CVDP and ChipBench, I keep coming back to one question: how far can AI go toward designing a complete chip on its own?
Today, we’re thrilled to introduce Build-A-Chip-Bench, the industry’s first AI benchmark for end-to-end chip design, built by ChipAgents in collaboration with @NVIDIA. It tests whether AI can take a design from specification all the way to verified GDS.
The top model scored just 16.6%.
Today’s models can produce functionally correct RTL in some cases, but reliably meeting physical requirements remains a major challenge. We still have a long way to go, and that’s what makes this research so exciting to me.
Proud of our research team and grateful to our industry partners for helping us give the community an honest measure of where we are and how much work lies ahead.
We have a group of AI founders from Australia coming to San Francisco this week as part of the @Latitude37AU immersion program and let me just say there is a genuine Aussie mafia in Bay Area tech scene and they are quite impressive and are turning out for this group.
Going to be fun!
Seeing a lot of AI for science companies pop up. I’m pretty bullish on the whole category.
Even the most pessimistic take on foundation models is compelling, they have basically the entire corpus of existing science and are really good at reasoning / combinatorials. A lot of “new science” is ultimately finding new ways to put existing building blocks together. OAI’s latest math results are a pretty good example of this already happening.
I also think it’s interesting to ask which fields get “solved” first. I think consensus is it's the ones with tightest closed-loop verification. Math has Lean. The easier it is to know objectively that you found something correct, the faster this should compound.
Airwallex is running its first Agentic Banking Hackathon and @t0finance is joining in.
We bring the financial context. Airwallex brings the rails to do something with it.
Pretty good setup for building some weird + useful stuff imo
Apply: https://t.co/EflS7WReBt
📣 Applications for the @Airwallex Agentic Banking Hackathon are now open.
If you're building agents that can move money, come build with us.
🤖 Airwallex APIs
💰 $100K in prizes & credits
🌉 Demo day in SF
Apply now: https://t.co/KTyhPfm7KS
Big thanks to the partners putting up prizes, credits and tools for builders: @visa, @claudeai, @awscloud, @altaxyz, @t0finance, @Base44 and @coinbase.
Among SMBs on @airwallex (reminder we cover markets around the world not just US) those paying for AI tools in Q3'25 grew processed volume 54% over the next year.
Non-adopters: 29%. They also trade with 2.7x more countries (8 vs 3).
Correlation, not causation. But AI may turn out to be a growth story, not just a cost story.
Airwallex is going BIG at #SFTechWeek - Dinner on the Giants' field; Private plane tour over SF; McLaren rides; and Agents, Automation & Ambition Panel discussion.
RSVP your spot 🙏
→ 10/5, Monday: Dinner on the Giants' field with baseball's home run king.
https://t.co/dDHfmdeuwe
→ 10/6, Tuesday: Agents, Automation & Ambition Panel
https://t.co/03Y6eEFjr2
→ 10/7, Wednesday: Private plane tour over SF
https://t.co/REmeloydJo
→ 10/8, Thursday: McLaren rides
https://t.co/oJgDTsjQi9
Spots are limited.