Robert Ambrogi Blog
The legal data company Entegrata, whose data lakehouse platform helps law firms consolidate siloed data into a single source of truth, has hired Andrew Baker, who for more than five years led applied AI and data science at Simpson Thacher & Bartlett, as its first chief AI and data strategy officer, where he will lead development of a new product line focused on AI enablement for law firms.
The appointment, announced today, comes at a time of rapid growth for the company, founder and CEO Tom Baldwin told me during an interview this week with him and Baker. “It took us three years to sign our first 10 customers,” Baldwin said. “It’s taken six months to sign the last 10.”
Among those recent signings are some of the largest firms in the world, including Cleary Gottlieb, Mayer Brown, Wilson Sonsini, Mintz, Kilpatrick Townsend, and Winston Taylor.
As I reported last year when Entegrata raised $4.5 million in seed funding, Baldwin — the former CIO or CKO at firms including Cadwalader, Reed Smith and Sheppard Mullin — founded the company in 2023 to address the fragmentation of law firm data across dozens of business systems, consolidating it within a purpose-built data lakehouse.
With Baker’s hire, the company is extending that mission from business intelligence into the data layer that underpins firm-wide AI.
A Reunion Two Decades in the Making
For Baldwin and Baker, the hire is something of a reunion. The two worked together at Reed Smith in the early 2010s, when Baldwin was running pricing, legal project management and other functions there and, as he tells it, went looking for “a unicorn in the form of a JD black belt” in process improvement, and found Baker’s profile on LinkedIn.
Baker later returned to Seyfarth Shaw, where he had previously worked, to help launch that firm’s innovation program. There, he founded what Entegrata says was the industry’s first legal technology innovations office and pioneered the legal solutions architect role.
He went on to co-lead the legal transformation and innovation practice at HBR Consulting (now Harbor) before joining Simpson Thacher, where he has spent the last five-and-a-half years advancing the firm’s use of data science and emerging technology. Along the way, he earned a master’s degree in applied data science.
“I’ve been hunkered down at this intersection of law, data, tech and strategy for quite some time,” Baker said.
So why leave what he called “probably one of the cooler jobs in a law firm that you could ask for” to join a vendor?
“It’s really about the opportunity to build something big and lasting in the industry,” Baker said. “I spent a lot of time in pretty much every vendor tool that’s out there, a lot of time working with teams to build custom solutions within our firm, and I feel like I have a pretty clear understanding of where the hard edges are on each of those. And we keep running into the same challenges over and over.”
Solving those challenges from inside a single firm is daunting, he said, but it is a different ballgame from the outside.
“When you think about where you want to put your time and energy if you’re doing technical work in a law firm, it’s not on the plumbing and the data ecosystem,” he said. “It’s building the cool, exciting, creative things.”
What ‘AI Enablement’ Means
In his new role, Baker will own Entegrata’s AI roadmap across engineering, product and customer experience, with his central focus on a new product line the company describes as AI enablement.
To date, Baker said, Entegrata has been adjacent to the AI story — its data could be put to creative use, but AI was one step removed from the core product.
“This next chapter in the book brings Entegrata straight to the coalface on everything AI-related,” he said. Whatever firms are pursuing — model routing, autonomous agents, custom development — “none of those can be executed very well without the right kind of content or context to make that happen.”
The goal, he said, is “providing a framework and layer to help people truly do that, and moving beyond theory into pragmatism.”
Baldwin says the product vision centers around what he calls the four Cs: content, context, cost control and choice. The idea is to give AI tools richer content and context drawn from across a firm’s systems, provide firms observability into AI costs, and preserve their ability to swap in new models and tools as the market shifts, without starting over each time.
“To be clear, we are not building yet another AI model or product,” Baldwin said. “We’re enabling deeper context and giving more content to the existing tools firms have” — he named Harvey, Legora and Deep Judge as examples — “and creating that downstream off-ramp to different tools as they become available.”
Central to the effort is an ontology for legal data that the company is currently building, layered on top of a schema that Baldwin said already contains some 5,000 fields. (When I wrote about the company last year, that schema had 3,500 fields.)
The ontology is designed to capture the full web of relationships within a firm — the “360-degree view” that connects a document to a timekeeper, the timekeeper to a matter, the matter to a client, and outward to industry, jurisdiction, judge and opposing counsel.
“You lose all of that context the minute you take a document out of the DMS and manually upload it into an AI tool,” Baldwin said. “We bring that all back and throw an ontology on top of it.”
He drew a comparison to the much-discussed partnership between Kirkland & Ellis and Palantir. “Long term, this is going to be an offering that’s going to feel very Kirkland-esque in terms of the capabilities, but at a fraction of the time and cost that it’ll take to implement something like Palantir.”
The Data Layer as the Bottleneck
Both men agree that the data layer has become the choke point for firms trying to move beyond their first wave of generative AI deployments.
In the early rush, Baldwin said, firms were under enormous pressure simply to get AI up and running, in part so they could say they were using it.
Now that firms have had time to settle into it, he said, the next stage of maturity is “moving beyond the disconnected download-documents-out-of-the-DM, upload-them-manually-into-an-AI-tool” workflow toward AI that can run on the firm’s full data estate. He said the last six firms Entegrata signed all characterized their moves as being about AI readiness and enablement.
The challenge for law firms, Baldwin said, differs from other industries. “The challenge isn’t in depth and speed. We don’t have petabytes of real-time streaming from one or two ERP sources. Our challenge is width. It’s stitching together data from 20 or 30 systems.”
Baker made a similar point about leading practice-of-law AI platforms. “The Harveys and Legoras of the world are fantastic for helping firms move quickly,” he said. “The challenge you run into is that they all have hard edges.” Firms whose practices do not fit neatly within those boundaries — or that want to rethink a practice’s economics or go-to-market approach — quickly find themselves breaking through them.
“Right now, there are a number of solutions that are focused on retrieval, and some of them do a really good job,” Baker said. “But that next hop over — how do we analyze, synthesize and create anew — it’s hard to get the data in the right structure for that. And we think we can do that in a way that others really can’t.”
As an example of what this can mean for customers, Baldwin said that for one of its largest customers, a 2,000-lawyer firm, Entegrata is building an AI tool for just three people in firm leadership designed to answer the questions that, as he put it, the people who sign everyone’s checks care about.
“It’s the opposite of, ‘I’m a third-year associate grinding through this asset purchase agreement draft,'” he said. “It’s, ‘I’m the chair of the firm and I want to know directionally where the firm’s headed.'”
Baker’s Remit
Baldwin described Baker’s mandate as spanning three buckets. First and foremost is architecting the new AI enablement product line. Second, with law firms increasingly hiring for net-new data and AI leadership roles, Baker will offer his playbook as a service to firms just standing up applied AI functions — drawing on his consulting experience at Harbor in making what works at one firm portable to many. Third, he will take over much of the data strategy engagement work that Baldwin has been leading himself.
“Andrew has lived this problem from inside one of the most prestigious firms on the planet, so he knows what it takes to make AI dependable across a real practice,” Baldwin said.
“His mandate is simple: prepare firm data and context so any model or tool it chooses actually delivers. That is the layer most of the industry is still missing, and it is what Andrew will build for our clients.”
Baker will be at ILTACON in Nashville, Aug. 23-27, where he hopes to meet with firms to discuss the intersection of data strategy and AI.