Legal Tech's Step-Function Moment: Why Legora Bet on the Last 20%

Max opens by noting it has been about 10 months since his last appearance, a period he treats as an eternity in Legora time. He describes telling Legora's CFO that he would sign the Series D term sheet before his Friday flight home from San Francisco. The term sheet arrived Thursday at 2 a.m., was checked by Legora and its lawyers, and was signed before he boarded.

He says the last 10 months have been astronomical: since October 2024 Legora has doubled the business every quarter, launched in the U.S. in March, and grown the New York team to 100. The company raised $550M, which Max believes is the largest round ever in legal tech, and saw more than $1.5B of demand. The New York team started during Legal Week with reps thrown into the deep end, including a laptop from Best Buy and a demo on the floor. That team has since moved into a base at Union Square.

In the internal announcement, Max told the team that raising a lot of money does not equate to building the best product. He said the funding is validation but does not help customers at all, and that everything Legora has done can be washed away. He argued the company must re-inject energy because the products built in 2023, 2024, and 2025 do not really exist in the product anymore.

Max argues Legora cannot build only for where it wants to be in 12 to 24 months because customers are not ready to meet it there. Instead, the company builds a staircase of features and products and teaches customers how to rethink their businesses along the way. He says Legora is in the business of selling the transformation these organizations and teams must undergo.

A concrete example is a 2024 feature that automatically flagged risks: Legora built an extensive taxonomy, categorized risks, and created a custom-tuned LLM flow to assign a risk to everything. Max says you can now throw a latest model such as Opus 4.5 or 4.6 at the problem and it solves it magically. That was a good product in 2024 and helped Legora get to where it is, but it illustrates why the team must have low ego. People need to work on something important for nine months and accept that it may then be washed away.

Max references a spider chart from Anthropic showing which tasks AI can do; once a task falls inside that realm, he treats it as conquered. The reality, he says, is that only about 1% of users have automated that particular task, so there is a big job in bringing the other 99% up to speed. That gap between capability and adoption is where Legora still has to do organizational work.

Max calls the customer experience a real step function: in summer 2024 Legora was about 60% accurate on limited-partnership-agreement key-term review, and by the end of that summer it was 100% accurate. The client then changed its business model, moving from an associate spending three days on a billable-hour model to a five-hour turnaround with manual checking and a fixed fee. Once a task can be done, Max says, it is conquered.

Max explains the law-firm predicament: if one firm offers venture financing at $250,000 or $100,000 and another offers it at $75,000 with twice the turnaround speed because of AI, the low-differentiation equilibrium breaks. Once a firm announces it has procured Legora and is more effective internally, competitors have to respond. More recently, enterprises such as banks, insurers, and pharmaceutical companies have woken up and are demanding that their panel firms leverage the technology. Law firms might prefer the old billable-hour model, but client pressure is changing behavior.

Adoption is more partner-dependent than firm-dependent because Max describes every law firm as basically 300 CEOs running their own books of business. Consolidating new ways of working is therefore hard. Forward-leaning firms use Legora for scale; one firm announced a Star Portal on Legora so clients can collaborate with it, and partners think about serving more clients with the same 14 to 16 hours a day. Legora gives them scale and a way to encode their IP and knowledge.

Max is convinced there will be a “firm AI” that leverages precedent, organizational memory, and context for how a firm does things. He points to the new paradigm of skills that models can download and tap into, and says practitioners will not be limited by systems that can do work. The limit will be human: how many agents can a person effectively guide? He compares running five Claude Code sessions at once to StarCraft, where mental load caps how many units you can control toward a single mission.

Inside Legora, Max pushes leaders and individual contributors to identify which parts of their work now fall under what models can do. Legora plans to scale from 400 to 900 people this year, but Max is not sure it needs to be that big to become a $1B ARR company if it leans into AI. He says superusers will become exponentially more valuable; a leaderboard of Claude Code and Cursor usage showed a real power graph, and while engineering has moved furthest, marketing, finance, and HR are next.

The hosts note that most AI software still sells seats, and Max says outcome-based pricing is extraordinarily difficult in legal work. An M&A deal in Canada differs from one in the U.S., India, or Australia, with local pricing power and a wide range of deal sizes from Paramount-scale transactions to small tuck-ins. He has optimized for speed to market: if buyers are used to buying seats, seats are the easier way to get up and running today.

Agentic launches change the economics because models do more work on the user's behalf and consume many more tokens behind the scenes. Max describes the models questioning themselves, double-checking, planning, executing, reflecting, and updating memory. That is when a platform may need to rethink how tokens are transacted, though he calls the answer a bit TBD.

When Anthropic launched legal skills and Thomson Reuters stock fell, Legora was about to start fundraising; public SaaS sold off while demand for Legora's private round exceeded $1.5B. Max interprets the reaction as a market correction: traditional SaaS companies without the velocity, talent, rigor, and ambition to claim the new green field should be less valued. He says models becoming generally useful helps Legora because the spotlight shows they can do more than write code.

Max argues that even if models do 80% of the work out of the box, the last 20% takes 99% of the time and requires taste, maintainability, and a team and vision for what comes next. Asked whether a “Claude legal” could substitute, he says “bring it on” and insists Legora must be the global category leader by investing beyond what models naturally do. He uses a database/AWS analogy: a database company that offered only what AWS later offered was toast, but AWS did not build everything, and the same will be true for Anthropic, OpenAI, Gemini, and Google. Legora has deliberately stayed a software provider rather than a service provider, and Max questions whether an AI-native law firm could beat incumbents like Kirkland or Goodwin that already have distribution.

Max tells executives from fast-growing SaaS companies that even strong companies like Klarna, Vanta, or Braze scaled slower than Legora. What those companies did in a year, Legora now has to do in a quarter. That requires leaders to lead from the front on AI, because you cannot expect the team to optimize its workflows if you are not doing it yourself.

The hosts point to Legora's bias to action, including a renaming decision made in about 24 hours. After the fundraise, Max plans to return to Stockholm and plant himself in the middle of the organization. Half the team has less than three months' tenure and lacks context on the two or three years that got Legora here, so culture carriers from the first two years must be vocal and visible.

Legora now serves a double-digit number of the top 50 law firms and some top 10 banking institutions, which expect a different enterprise software cadence. Max says CIOs and CTOs often optimize for “safe” and their own comfort when they should optimize for speed. He asks clients to be willing to drop what worked for what will get them to the next stage, and he says Legora only wins if its clients win.

Max describes a 75-year-old partner asking whether he should lean into AI or retire; after seeing use cases, the partner called it the coolest thing he had ever seen. Max told him that is the energy the entire firm needs to bring. Legora's pitch is that it has done 800 deployments and knows what works and what does not, so it will make decisions based on a worldview and invite customers into the “tent.”

Max describes the tent as a place of honesty: Legora will say what is working, what is not, and what it wants permission to do. He says many buyers have never met a more ambitious team, and that the mission is not selling legal technology but transitioning entire teams to be AI-native. He calls it the most exciting work transformation since the web.

Legora hires people who are collaborative and competitive, willing to fight for clients and help them get to the next stage. Its legal engineers are ex-practitioners who act as the glue and “buddies” for new customers, having felt the pain of legal work firsthand and now reinventing it with firms. They also get to elevate their own careers by combining practice knowledge with technology.

Max gives executives leeway to run their functions, but in the weekly executive meeting everyone is “team Legora” and optimizes globally rather than representing marketing or engineering. He values a mix of executives who have seen scale and AI-native, high-IQ people who question authority. If you question someone enough, he says, they start to realize why something does not make sense, and a high-trust room makes things easier.

Max says December was a big shift when the team played with Opus 4.5 and 4.6. He describes telling an agent to do something, forgetting to add context, and the agent asking for the missing file; the internal thinking even reasons about why Max did not provide it. The interaction feels more human and less friction, and the improvement is decreasing at an exponential pace.

Once you have had an agent experience like the missing-file interaction, Max says the challenge is bringing that capability into your own domain. Legora wants to be AI-native across the whole stack so inspiration from engineering or marketing can flow into legal. In software engineering an agent can open a pull request, but in professional services the UX of collaborating with agents on matters is less clear, so more of the system around agents must be built.

Max says balancing continuity against whipsaw does not have a binary answer. Legora needs a resilient, low-ego team willing to accept the new world and run at an intensity that is practically unheard of. He contrasts Google-style bottom-up empowerment, which works with infinite distribution, with Legora's need to be crisp about what it is building and get the whole company rowing at the same time.

Max adds a nuance to product development: are you building for the CIO or partner who buys, or for the end user? In 2024 and early 2025 Legora focused on getting in the door and refining foundational building blocks, similar to Ramp having to show up better than Brex. He warns against getting stuck in incrementalism while still valuing iteration.

He distinguishes iterating from incremental improvements: he optimizes for iteration cycles and then inserts bold ideas when a threshold is crossed. Legora has an R&D function, and engineers pushing a feature can also explore a crazy idea. Legora finds out what new models unlock through many evals and by hiring lawyers who work with product and engineering; 70 to 100 legal engineers are the most savvy users. In capital markets, pairing an engineer and a lawyer for a week produced amazing results for about $500 in model runs, and the next question is whether it can be generalized across capital markets, finance, M&A, litigation, and in-house work.

Max says the round supports an all-in U.S. push: Legora opened New York in March, then Denver, Chicago, and Houston, with more hubs planned this year. The strategy follows legal work's central hubs, and because Legora started in Sweden—whose entire legal market is smaller than Kirkland & Ellis—it had to internationalize across jurisdictions early.

In the U.S. Legora is now working with big law firms, private equity firms, pharmaceutical companies, and banks, creating a network flywheel. Max wants legal work to be synonymous with Legora the way design work is synonymous with Figma. He jokes about the Swedish phrase “blood smack,” which became an internal hashtag after a reporter translated it as a metallic taste of blood.

The U.S. expansion had a rule: do not open and hire until Legora could serve two of the best firms in the market from Sweden and London. Cleary Gottlieb and Goodwin Procter started working with Legora, giving the team confidence to open the office and go hard. Max says the most important thing was creating “one Legora feeling” across offices; everyone in New York goes to Stockholm for onboarding, and Legora is explicit about five days in office and dinner together every day.

Asked about Stockholm's ecosystem, Max cites technical talent, role models like Spotify, Skype, and Klarna, and an Americanized, pro-tech culture with a “how hard can it be” mentality. He also says he changed his mind about long-running agents: tools that can work for a long period and finish accurately arrived sooner than he expected. That has made him want Legora to move faster into a world where agents do work on users' behalf.