Max is the founder and CEO of Lagora, a legal AI company that has grown to nearly 400 employees in just two years. Chatham Gujral is a partner at Benchmark who led Lagora's seed round in March 2024 and sits on the board. Jack is the host conducting this conversation about competition, culture, and building AI-native companies.

The discussion covers Lagora's journey from a struggling 2020 startup working with racist Swedish BERT models to becoming a dominant legal AI platform after GPT-3.5's release. Max explains how the founding team embedded themselves in Stockholm law firms to understand the industry deeply, while Chatham shares his investment thesis around legal market dynamics and AI adoption patterns.

Key topics include Lagora's unique Stockholm-centric culture where employees 'taste blood' from working intensely, their strategy of focusing on just three core features to achieve rapid growth, and their approach to building AI-native organizations that can quickly adapt as model capabilities evolve.

From Swedish BERT Failures to GPT-3.5 Success

Lagora's origins trace to a 2020 company with four co-founders working on AI and law intersection, initially using Swedish BERT models that were 'blatantly racist because it had been trained on Swedish forums' - Max

The breakthrough came when GPT-3.5 launched: 'When the LLMs like 3.5 came, that was when the moment shifted, right? And so we turned this into a company' - Max

Max joined as two co-founders left, and the team decided to 'run like hell' in the AI-law direction without knowing the exact product they'd build

Embedded Learning and Customer-First Development

The founding team embedded in a windowless Stockholm law firm conference room to understand legal workflows, while Max bought lawyers lunch on LinkedIn to learn about different practice areas

Their first lawyer hire was actually a customer - the CIO of a major Swedish firm who had built his own GPT-document system and decided 'these guys are going to run faster than me' - Max

Chatham's investment thesis centered on Max's clarity that general foundation models would serve legal data uniquely well, contrary to the 2023-2024 paradigm of training custom models

The Three-Feature Focus That Doubled Revenue

In September 2024, facing 15 possible features to build, the entire 10-person company gathered for wings, beer, and peanuts to decide on just three core capabilities

The chosen features were tabular extraction, deep Word/Outlook integration, and embedding Lagora wherever lawyers already work - leading to revenue doubling from $1.5M to $4M in one quarter

Max wrote a 'very short product manifesto' that rallied the company around being best-in-class at these three things rather than mediocre at fifteen

Legal Market Dynamics and Competitive Equilibrium

Law firms operate in 'perfect equilibrium with frankly like pretty low differentiation' - if one firm adopts better AI tools for faster, cheaper service, all competitors must follow - Chatham

Legal was 'so underserved with great software for such a long time that there was this like a lot of built-up problems that we could easily solve with LLMs' - Max

Lawyers are 'extremely well educated' and 'tech savvy' - they were already using ChatGPT and Claude, so legal AI products had to be demonstrably better than foundation models - Chatham

AI-Native Organization Design and Culture

Over 10% of Lagora's engineering team are ex-YC founders, including head of engineering Jake and VP product Adrian, creating a culture of treating each feature like 'your own company'

The company operates without long-term roadmaps, sometimes changing direction daily as model capabilities evolve: 'Every upgrade was pretty incremental. But now it like flipped. Opus 4.6 flipped in capabilities' - Max

Max tells every executive they're 'joining with an expiration date' and must continuously prove they can scale with the exponentially growing business

Stockholm Culture and Global Expansion Strategy

All employees globally must onboard in Stockholm and interview there, with culture carriers from Sweden seeding international offices to maintain consistency

The company serves dinner at 8 PM daily across all offices, stemming from Max's McKinsey internship experience, creating a unified intense work culture

Max's Swedish interview about working so hard 'you taste blood' was mistranslated to English as 'we wake up with a metallic taste of blood in our mouths,' becoming an internal meme

Evaluation Infrastructure and Model Capabilities

Lagora built proprietary eval infrastructure that can identify latent model capabilities unknown to foundation model companies themselves, with customers contributing manual tasks as evaluation benchmarks

A Danish law firm's LPA key term review task went from 60% accuracy in summer 2024 to 100% by end of summer - once conquered, 'that task is done. Like it's over' - Max

At Anthropic's customer advisory board, the consensus was that models are no longer the bottleneck - it's the software infrastructure around putting models in trustworthy execution environments

Series D Fundraising and Future Vision

The Series D round generated $1.5 billion in demand, with Excel leading and participation from Menlo and Bain Capital

New CFO David from Vanta chose to walk out to Kanye's 'Monster' at the company kickoff, earning the internal title 'CFGO' (Chief Financial Growth Officer)

Max and Chatham's original equity negotiation was conducted 'decimal by decimal' in Excel until reaching 19.521% ownership, with both parties 'equally unhappy or happy'