
Lead sales @harvey. Love pipe gen, hate complacency.
New York
Joined January 2010
- Tweets26
- Following1.1K
- Followers2.1K
- Likes19
Rob Saliterman retweeted
We’re not aware of any law firm going to this extreme.
For example Latham is buying GPUs but is also firm-wide on Harvey and we announced a deeper partnership today.
We think that’s the right strategy and what matters more than build vs buy is how heavily a firm is investing in AI.
The best way to invest is a portfolio approach of working closely with vertical vendors and the labs and cloud providers while also self-building with open source and the frontier ecosystem.
Everything is moving too fast and law firms are too complex for a single binary approach like this post suggests.
Law firms will not rely on a single vendor nor will they build the entire stack themselves.
My view is that longer term the labs and clouds will build general intelligence infrastructure and vertical players will help customize it for each industry.
Then individual firms will leverage their unique data, human expertise and workflows to build differentiated AI to better compete against their peers and serve their clients.
Our goal is to build the trusted vertical layer and open it up through APIs so firms can build everything from custom models and context layers to agents and products with us.
And most importantly do it in a way they can trust that their data and their clients’ data is safe and only used for their benefit.
Rob Saliterman retweeted
Harvey is hiring a government affairs and public policy leader. The role: engage policymakers in the US and around the world so AI strengthens the legal profession and expands access to justice. You'd build the program working directly with me.
harvey.ai/company/careers/c5…
Pretty good for a company that’s not even four years old! Amazing what you can do with maniacal customer focus and elite talent density. @winstonweinberg @gabepereyra and the whole team are absolute beasts.
Q2 recap for @harvey
- +$100M NNARR
- 53% DAU/MAU
Key hires (including Q1)
- Anique (CPO) - prev VP of Product at Rippling
- Rachel (CMO) - prev CMO at Notion
- Brooks (CISO) - prev CISO at Roblox
- Keith (CSO) - prev CPO at Google
Product
- Agent unification - cloud agents can use all Harvey product surfaces
- Command center (EA) - monitor adoption and ROI by use case
- Contract intelligence (EA) - agentic contracting platform for enterprises
Eng
- Migration to cloud agent infrastructure
- Integrating open source inference providers
- Scaling document processing (54TB / week)
AI
- Legal Agent Bench
- Open source post training
- Published multiple research directions with partners
We invested heavily in cloud agent infrastructure at the end of last year and in Q1. In Q2 we also unified many of our product surfaces (collapsed as @winstonweinberg says) by making them all tools accessible by our cloud agents.
Prior to this, there were a lot of capabilities in Harvey that were often only discovered by power users. As cloud agents get better and our product becomes more connected we are seeing users discover more of the product by learning from their agents (see plot of product surfaces per user).
Rob Saliterman retweeted
Super valid question on why we are working with multiple research partners.
Our goal is to help the entire legal industry build better models, not just ourselves. This is why we open sourced our legal agent benchmark and the more people use it the more we learn how to make it better for everyone.
Post training is still an open field of research and there isn’t a one-sized fits all solution. We have a huge amount of open research problems, product surfaces and task types that may all require different post training approaches. Each partner explored something different:
1. @baseten - novel compaction strategies for understanding large corpuses like datarooms.
2. @FireworksAI_HQ - model routing and frontier-as-advisor-tool for cost reduction.
3. @trajectorylabs - continual learning with nemotron.
4. @EngramLab - learning firm knowledge / enterprise and precedent search.
5. @appliedcompute - harness and compaction optimization on AC2
We have many more research directions we want to explore. On top of that we need to ensure these work with many different types of open source base models. Some customers want @Zai_org or @Kimi_Moonshot, some want a European model like @MistralAI and some want a US model like @nvidia or @reflection_ai.
I think the big misconception is that the goal is to train a single model and then you're done. What every company and industry needs to build an ecosystem and system for constantly improving models which requires data, research, training, etc to be shared so everyone can contribute and benefit.
Our hope is that by open sourcing our dataset and showcasing the amazing work of all these partners we will encourage more talented AI folks to focus on legal and help the industry build better AI systems overall.
Rob Saliterman retweeted
Our new Congressional District targeting in action: @lowbrowkate covers how the tool works for reaching voters. goo.gl/lBuhG
Rob Saliterman retweeted
Voters now use 4 screens to access political info like never before. Learn more in our post/infographic: goo.gl/hlVoi
Rob Saliterman retweeted
Happy Birthday to our friend Tony Snow.. he would have been 57 today. We miss him!
Rob Saliterman retweeted
Millions of Americans Have Cut the TV Cord [STUDY] - on.mash.to/HPk0xE
Rob Saliterman retweeted
See what is happening on the ground in New Hampshire google.com/elections/ed/us/o… #FITN





















