Our strongest quarter yet, alongside the launch of Max, Matrix 2.0, and our API and MCP server.
Thank you to our customers for your continued trust and to the Hebbia team for the work behind these results.
Q3 was the biggest quarter in Hebbia's history, and I’m incredibly proud of the team that made it happen.
• New ARR tripled quarter over quarter
• Daily active users grew 5x
• Gross revenue retention exceeded 95%
The usage growth is especially exciting to me. More people are using Hebbia, and a greater share are coming back every day.
We also shipped three major launches:
• Max, an AI teammate that works the way your firm works.
• Matrix 2.0, our next generation of agents for research and analysis.
• Our API and MCP server, bringing Hebbia into the tools our customers already use and build.
Seeing the team deliver these products while more customers make Hebbia part of their daily work makes this quarter particularly meaningful. Last week alone, we won five new banks to kick off the quarter.
Thank you to our customers for trusting us with your work, and to everyone at Hebbia for making this our strongest quarter yet.
Hebbia now partners with @Databricks. Deal teams can now bring their firm’s own data into Hebbia, whether that’s portfolio metrics, fund performance, or any dataset in Databricks.
Link in comments.
2
11
876
Today, we’re launching knowledge graphs in Hebbia, so financial institutions can connect what they already have and act faster on what they find.
Introducing Knowledge Graphs
Hebbia models the context inside your firm so every person and agent can make better decisions from it.
Hebbia retweeted
revenue per employee tells you a lot about company culture & talent density
jane street's revenue/employee only rivaled by anthropic
ramp is highest among fintech startups, but still lower than goldman
Hebbia retweeted
design.hebbia.com/brand !!!
walk with me... most modern institutional brands are 90% random stock images of skyscrapers.
they either don’t have an owner of taste, or are afraid their audiences will be turned off by anything too opinionated.
we’ve spent the last year challenging that assumption. studying the entanglement between finance, culture, and the arts.
talking to many leaders in finance, we have come to understand that choosing a career in finance often comes with an aspiration to be surrounded by the highest craft: architecture, fashion, food, products, experiences, etc.
we set out to build an identity that celebrates this pursuit. we wanted to respect the history of the institution, while being technology forward and side-stepping nostalgia slop.
our research was mostly offline: visiting print magazine archives, hand-scanning buildings in fidi for material & texture, & talking to customers about design that motivates them.
i think the work speaks for itself.
We’re adding our Financial Services Benchmark to @FireworksAI_HQ’s Specialized Intelligence Index (SII).
Joining the SII places our evals in a broader comparison of model capabilities, so investing and banking teams can choose the most useful model for the work at hand.
4
13
1,347
Hebbia retweeted
Honored to be on UBS’s list of the top 10 private companies to keep an eye out for.
In great company alongside: @databricks, @anduriltech, @Replit, and @perplexity_ai ...
Hebbia retweeted
Great conversations in Bordeaux with Chief Investment Officers from the world’s 50 largest insurers, including many @hebbia customers.
AI momentum is accelerating across Europe. Excited to share lessons from working with the world’s leading asset managers.
Thanks to Jean-Baptiste Tricot, @AXA , and the @TheGenevaAssoc for bringing us together, and Martin Zingg and Patrizio Urciuoli for a great discussion on the CIO’s AI toolkit.
Hebbia is partnering with @DatasiteGlobal to bring live data room content into Hebbia.
Now customers can analyze Datasite’s data room content alongside their firm’s internal deal context and market data, with every answer cited to its source.
Read more in the comments.
Proud to partner with @AfterQuery and @nvidia to build more efficient models for investors with our credit expertise.
Replying to @AfterQuery
@AfterQuery just moved an open model past the cost-accuracy frontier in finance.
We post-trained @NVIDIAAI Nemotron 3.5 Super VL on expert credit tasks with @hebbia.
Perfect solved: 47% to 74%. That beats Sol 6, and lands within 3 points of Astra 6, at about 1/7 the cost.
Link to the full case study in replies.














