No Priors Podcast: Scaling Legal AI and Building Next-Generation Law Firms

Guest: Gabe Pereyra (Co-Founder & President, Harvey)
Hosts: Sarah Guo & Elad Gil
Date: December 5, 2025

Episode Overview

This episode explores the transformation of the legal industry through AI with Gabe Pereyra, co-founder and president of Harvey, a fast-growing AI platform for law firms and large enterprises. The discussion ranges from the origin and growth of Harvey, the particular challenges and opportunities in legal workflows, technical analogies between law and software engineering, how AI is changing law firm business models, and Gabe’s unique perspective as a founder with a research background.

Key Discussion Points & Insights

1. What is Harvey & Its Scale

• [00:15] Gabe: Harvey builds AI software for law firms and in-house legal teams.
  • Nearly 1,000 customers, 500 employees.
  • Rapid scaling over three and a half years.
  • Early investment from OpenAI; began by giving GPT-4 to lawyers and learning from their interactions.
• Purpose Shift: From productivity for individual lawyers to team- and firm-wide productivity, emphasizing "orchestration, governance, and all of the enterprise product problems that you run into at scale."
  • Quote: “The big problem we're solving is not how do you make individual lawyers more productive, it's how do you make a team of lawyers… more productive? And... how do you make an entire law firm... more profitable?” — Gabe Pereyra [01:08]

2. Product Differentiation vs. Copilot/ChatGPT

• Harvey started as a “prompt a model” tool, but quickly encountered the need for context-awareness, reliability, and workflow integration—unique challenges in law.
• Quote: “The industry was so text heavy... you got so much value from just interacting with the models. And then... you ran into all the sharp edges... they hallucinate, they're not connected to a bunch of our context.” — Gabe Pereyra [00:43]

3. Expanding Beyond Law Firms

• Now serving Fortune 500, large PE, and Global 2000 clients, offering a platform for in-house legal operations and secure collaboration between law firms and their clients.
  • Example: Signed Walmart and AT&T as enterprise clients.
• Major technical challenges: security and data privacy in collaborative workflows.
  • Quote: “What we're starting to build is a platform for the in-house teams... [and] the collaborative tissue… securely share this data with my law firm.” — Gabe Pereyra [02:14]

4. Legal Workflows Explained

• Legal work, especially for large firms, is vastly more complex than consumer legal matters (e.g., leases).
• Example: Fund formation for PE/VC involves complex entity structures, negotiation, and project management akin to programming in an unfamiliar codebase.
  • Quote: “One analogy you can think of is understanding a code base, but the code base is all these contracts and all this legal work.” — Gabe Pereyra [05:26]

5. Agentic AI & Legal Workflows

• Harvey is starting to implement “agentic” workflows—AI agents executing sequences of tasks with oversight, mirroring associate workflows.
  • Associates are likened to agents who deconstruct high-level strategies into research and drafting tasks.
  • Drawing on past DeepMind RL research for these concepts.
  • Quote: “You can kind of think of associates as agents... they get this task from a partner... ‘Can you go research that, look it up, cite it, write me a memo.’ And so a lot of the systems we're starting to build look a lot like that." — Gabe Pereyra [06:46]

6. Law Firm Structure in an AI World

• AI challenges traditional “leverage models” (100 associates:10 partners:1 eventual partner).
  • Law firms may need fewer associates, which could affect partner pipeline and training.
  • AI accelerates learning for junior lawyers, akin to how LLMs changed learning programming.
  • Quote: “The role of law firm partners actually doesn't change that much. In the same way, I don't think the role of very senior engineers changes with this because you're largely delegating work and what you're getting paid to do is... the right abstractions... and interface with the client.” — Gabe Pereyra [12:50]

7. Cultivating Legal Expertise & the Gordon Moody Analogy

• Senior partners like Gordon Moody (ex-Wachtel) possess “reasoning traces” built from years of experience in unstructured, complex deals; this expertise is still beyond current AI capabilities.
  • Quote: “If you're building search at Google, these people can just point out, ‘Hey, at this scale, it's going to collapse for some reason that’s not intuitive.’ ...That technical understanding of how you architect these things... that's the process and a lot that’s a reasoning trace.” — Gabe Pereyra [14:16]

8. Challenges of RL and Evaluation in Law

• Unlike code (unit tests) or math, law is not easily verifiable—most legal evaluation is subjective and based on partner-level review.
  • Internal feedback data is critical for model training and evaluation.
  • Long-term outcomes (mergers that hold up, litigation avoided) are the “real” tests, not simple pass/fail checks.
  • Quote: "If you think of what that reward function is at the law firms, it’s the partners... There is no way to verify besides the senior partner who's done a bunch of these.” — Gabe Pereyra [17:41]

9. Enterprise Adoption & Implementation Playbook

• Harvey launched a “Deployed Engineering Force”—not full custom dev, but a technical team to help large clients integrate, connect data, and adapt workflows.
  • High demand for technical consultants to help map operations into generative AI.
• Law firms themselves are becoming Harvey implementation partners for their clients—a new revenue path.
• Quote: “We just want smart technical people to sit here and help us think about our business and our operations and how we should start mapping that into gen systems.” — Gabe Pereyra [20:12]

10. Transforming Legal Business Models

• Shift from mere individual productivity tool to driving firm-wide transformation and profitability.
  • Surprising speed of adoption by large, tradition-bound law firms.
  • Early bets paying off in text-heavy, knowledge industries.
  • Quote: “It was still surprising how quickly some of these law firms adopted this… This technology is so transformative for these industries that just are so text heavy and knowledge based.” — Gabe Pereyra [24:39]

11. Why Not Build a Law Firm?

• Core insight: Don’t compete with customers. Harvey’s success is in enabling all law firms to become “AI-first,” not in becoming a law firm itself.
• Building a law firm and a tech company concurrently is too complex; better to empower, not replace.
  • Quote: “You're essentially just building two different companies, right? …The best outcome is if we can figure out, how do we make every law firm… become an AI first law firm, not how do we build one ourselves?” — Gabe Pereyra [25:39]

12. Ecosystem & Market Scope

• Legal is a trillion-dollar market; professional services $3–5T, all requiring secure, collaborative, AI-infused workflows.
  • Not just law firms, but banks, consultancies (PwC), HR firms—platform approach is key.
  • Quote: “How do we build the platform that lets professional service providers and their clients collaborate?” — Gabe Pereyra [28:21]

13. Founder Journey & Surprises

• Transitioning from AI research to CEO/founder required a major “mental model shift” to scaling people, orgs, and technology.
  • “I think the shift… was just how much I had to change my mental model of the type of company we're building. How you do this at scale, how you operate…” — Gabe Pereyra [29:51]
• Harvey started before GPT-4; early conviction in LLM capabilities (thanks to research experience and cofounder Winston’s legal insights).
  • Key moment: GPT-4 capabilities enabled use cases that 3.5 could not.

14. Product Mindset: Open-Ended Ambition

• Harvey always pursued the ambition to handle all legal work, not just simple tasks, paralleling how code LLM products evolved.
• The importance of giving users tools that empower complex, unstructured workflows—like “upload a document and do something with it.”
  • Quote: “These models can help you do any programming task in any programming language… We felt that same way in legal.” — Gabe Pereyra [34:04]

15. Market Timing & Analogies to Coding/AI

• Discussed why legal LLM products scaled faster than coding ones—perhaps due to earlier focus on product form factor and instant utility in legal.
• Quote: “The initial feature we built that none of the products had at the time was upload a document and do something with it. Right. That is a lot of legal tasks.” — Gabe Pereyra [36:17]

16. Hiring & Growth

• Harvey is scaling: hiring engineers across frontend, product, AI; growing New York office.
  • “Anyone strong engineer, please apply.” — Gabe Pereyra [37:40]

17. Fun Personal Asides

• Founders’ fitness challenge (15 pull-ups in one set) [38:13]
• Gabe’s semi-ascetic founder lifestyle—story about not having a bed frame due to startup busyness [39:21]

18. AI’s Future: Organizational Transformation

• Gabe predicts most people (even in tech) underestimate how much better models will get and how rapidly.
• The next frontier: AI transforming not individual tasks, but whole organizations—unlocking new paradigms for collaborative productivity.
  • Quote: “A lot of the things we're starting to think about is organizational productivity and how do you build these systems at scale... making someone program 20% faster doesn't make you build a product 20% faster.” — Gabe Pereyra [41:20]

19. How Will AI and Humans Work Together Long-term?

• The real challenge is not “AI replacing everyone” but orchestrating human-AI collaboration for complex, unstructured tasks—analogous to how Figma changed design teams.
  • “How are humans and AIs going to work super effectively? ... it's like so much of it is how you organize all of these.” — Gabe Pereyra [42:49]

Notable Quotes

• "You can kind of think of associates as agents... Can you go research that, look it up, cite it, write me a memo. A lot of the systems we're starting to build look a lot like that." — Gabe Pereyra [06:46]

• "We're starting to connect to a lot of their business systems... there is just this massive amount of work where we go to a large bank and they say, we don't have any document management system for our legal department. Can you just build us one?" — Gabe Pereyra [20:12]

• "The role of law firm partners actually doesn't change that much. In the same way, I don't think the role of very senior engineers changes with this." — Gabe Pereyra [12:50]

• "If you think of what that reward function is at the law firms, it's the partners... there is no way to verify besides the senior partner who’s done a bunch of these." — Gabe Pereyra [17:41]

• "How do we build the platform that lets professional service providers and their clients collaborate?" — Gabe Pereyra [28:21]

• "A lot of the things we're starting to think about is organizational productivity and how do you build these systems at scale... making someone program 20% faster doesn't make you build a product 20% faster." — Gabe Pereyra [41:20]

Episode Timeline

| Timestamp | Segment | |:-------------:|:------------------------------------------------------------------------------| | 00:09-01:08 | Harvey’s scale, market focus, early product vision | | 02:14-03:21 | Enterprise focus, security/collaboration challenges | | 03:48-06:20 | Deep dive: legal workflows, contract/fund formation analogies | | 06:20-09:32 | Agentic AI in legal, RL analogies, associates as agents | | 09:41-12:26 | Law firm future, training partners, evolving org models | | 12:26-14:16 | Law partners vs. AI, expertise analogy (Gordon Moody) | | 17:41-19:46 | RL, model evaluation, legal vs. code verification challenges | | 20:12-23:11 | Harvey’s deployed engineering force, enterprise implementation | | 23:45-25:27 | Law firms as transformation partners, rapid customer adoption | | 25:39-29:25 | “Why not build a law firm?”—platform vs. vertical integration | | 29:51-31:27 | Founder perspective, research background, early conviction | | 34:04-36:00 | Product ambition, legal/coding analogies, keeping form factor open-ended | | 36:17-37:40 | Hiring/growth, scaling the team | | 39:21-40:17 | Fun Q&A: founder mattress story | | 41:12-43:24 | AI’s future: capability growth and organizational transformation |

Memorable Moments

• Gabe’s analogy of a lawyer’s workflow to navigating a codebase, highlighting both as complex, context-rich, and long unamenable to software tooling.
• Harvey’s origin story: The leap of faith on LLMs’ capabilities, not just for basic automation, but for the full complexity of legal work.
• Personal touch: Gabe’s bed frame story as a symbol of all-in founder commitment [39:21].
• Team fitness banter: Founders can do 15 pull-ups in a set [38:13].

Summary in One Sentence

Harvey’s journey shows how AI, when deeply integrated and tailored to complex professional workflows like law, can drive not just individual productivity but transformative organizational change—reshaping what is possible for entire industries.