Max Junestrand (Legora): Raising the bar in legal AI

Max Junestrand is the co-founder and CEO of Legora, the AI platform built for legal work. Max came to legal AI as an outsider. A former esports player and business school graduate, he joined two technical co-founders in 2023, when access to GPT-3.5 made a previously difficult problem suddenly solvable. In less than two years, Legora has scaled from serving a single law firm in Stockholm to having clients in more than 40 markets. Their team grew from 40 to 400 people in under 12 months, achieving 300% net revenue retention in 2025.

In this interview, Max explains how Legora built trust in a market that had already been burned by AI, what it takes to expand into a new country every few months, and why thinking small is its own kind of mistake.

This article is part of Perspectives, a series of conversations between Eléonore Crespo, co-founder and co-CEO of Pigment, and the technology leaders shaping how AI gets built, adopted, and scaled.

Law picked Legora before Legora picked law

Max's two co-founders, August Erséus and Sigge Labor, had been experimenting with legal AI since 2020, long before the current wave arrived. At the time, they were working with BERT-based models, including a Swedish-tuned version trained on Reddit. It was not, as Max puts it, very smart, and it had some additional problems that came with that training data.

Everything changed when they got access to GPT-3.5 in early 2023. They reached out to Max, who had been working in venture capital, because they needed someone with a business background to complete an application for a local accelerator. He was, in his own words, the “token business school guy” on the PowerPoint. But from that point on, none of them looked back.

What struck Max about the legal space from the start was how much of it remained unsolved. Legal work is detailed, high-stakes, and document-heavy – something early generative AI tools weren’t yet equipped to handle. With the new models, a huge surface area of challenging problems suddenly became buildable.

The first paying client was a friend's mom

Legora's first paying client found them at a party. A friend mentioned that his mother used ChatGPT all the time, and Max suggested she try their platform instead. She paid $100 a month and is still a client today. Max describes this as 100% cohort retention from July 2023.

The client that really opened the market, though, was Mannheimer Swartling, the largest law firm in the Nordics. Max managed to get into the managing partner's calendar – something he was told would be harder than breaking into Fort Knox – and showed up with a clear pitch: “You don't know AI. I don't know law. AI is going to be incredibly important for your business, so we should be working together.”

Because Legora couldn't afford office space on their angel round of €50,000, they ended up working from within the firm itself. That proximity accelerated everything. They passed security, compliance, and data privacy reviews, then brought the product directly to lawyers in the lunch canteen, asking them to input legal queries and see what came back. When a few use cases worked well, interest spread quickly through the firm.

The fact that the top firm in the market had bet on Legora meant other firms soon followed. That pattern – find the most credible partner in a new market, deliver for them, and let the market draw its own conclusions – became Legora's playbook for expansion.‍‍

Being a second mover meant arriving when the product could actually work

In 2023, many early legal AI companies made big bets on fine-tuning AI models. But Legora made a different call: use every frontier model and build all functionality like boats. That way, when the tide rises, everything improves.

Rather than trying to own the model layer, Legora focused on the application. How do you work with a hundred documents, not just ten? How do you connect to the web? They identified the spaces where the bigger players had gaps and put their energy into the platform.

As it turns out, timing matters. Max watched a number of early legal AI companies push products into the market before the models were ready to support them. Lawyers tried those tools in early 2023 and came away unimpressed. By the time Legora was in front of those same lawyers in mid-2024, with something that actually worked, they built more trust than they would have had they arrived first.

Legora also made a deliberate decision not to sell right after their Series A. It was not popular with their investors at the time, but it was the right call for the long run.

Agents are now doing the work that partners used to do themselves

In 2025, the primary users of Legora were associates – junior lawyers using the platform to work faster, get a second pair of eyes on documents, and handle volume. That made sense for an AI tool that was primarily augmenting individual work. By 2026, partners had started using it directly. Rather than delegating a task to an associate, they were giving it to an agent.

Partners are the most senior, most time-constrained, and most skeptical people in a law firm, so the shift towards partner adoption was a sign they could trust Legora’s output enough to stake their work on it. A third-party study of Legora's clients found that 71% of lawyers using the platform surfaced things they otherwise would not have found, like relevant precedents, overlooked clauses, and risks buried in data rooms. At that point, not using AI started to look like a professional oversight.

Legora generated 300% net revenue retention in 2025, and the growth isn’t coming from new logos alone. It’s coming from deeper embedding in the firms they already serve and from expansion into adjacent areas including tax, compliance, and risk.

In an enterprise setting, speed looks different

Legora plans product roadmaps on a quarterly basis, even though they can build faster than that. The reason why is their enterprise clients.

Enterprise clients need to know where the product will be in three months so they can plan their own implementation and adoption timelines. Even if Legora could ship something in two weeks, the lawyers who actually use it need time to change the way they work. 

The agentic tools Legora's engineering team uses internally have compressed their own build cycle dramatically. Bug reports from Slack get triaged into Linear automatically, Cursor looks at the issue and produces a first pull request, and a review bot assesses security risk before a human looks at it. Very little code is written manually. Max describes this as something that would have sounded impossible two years ago.

His goal is to take the habits that engineering, product, and design (EPD) has built and make them accessible across the rest of the company by building a central tooling hub that curious people in any function can pick up and run with. The teams that move fastest at Legora are the ones that have already internalized this.‍

European founders have to learn the motion of going to new markets

Legora started in Sweden and moved quickly to Finland, Spain, France, and Germany before making the jump to the US. Max is direct about why: individual European markets are too small to build the kind of company he wants to build.

The strategy for each new market has been consistent. First, the product needs to be integrated with local legal knowledge – case law, legislation, and other sources lawyers in that jurisdiction actually rely on. Second, Legora looks for a winning brand to partner with. When they went to India, for example, they partnered with one of the country's top law firms. When that partnership became public, every other firm in the market wanted to talk.

The product, it turned out, internationalized well from the start. The legal technology stack for large enterprises and law firms is remarkably similar across markets. That made it easier to adapt than Max had expected. The harder part was internal, going from 40 to 400 people in under 12 months, across multiple time zones, while also keeping the organization coherent.

Max spent the better part of 2025 on the road. That reality, he says, is just part of building globally from Europe.‍

Building from outside the dominant ecosystem has its own advantages

Max is proud of being a European founder building a global company. He is also clear-eyed about the structural challenge. Statistically, you are more likely to succeed starting in San Francisco, and certain talent gaps – particularly in enterprise B2B sales – are harder to fill from Stockholm.

But he believes European companies have something that is genuinely harder to build in the US. Legora's team are missionaries, not mercenaries. They came together around a specific problem and a specific ambition, and the shared experience of building it from outside the dominant tech ecosystem has bonded them in ways he thinks would be harder to replicate if everyone were cycling through their third or fourth flavor-of-the-month company.

The competitive chip matters too. Not wanting the West Coast to define the rules of the game is, for Max, a real source of motivation. Legora now has a large office in New York, and the American and European parts of the organization have started to blend. But the founding culture – curious, fast-moving, we'll just do it ourselves – is something he actively protects.

Hiring for the AI world also means hiring for curiosity and ambition

Max looks for three qualities in candidates: a demonstrated history of doing something exceptional, genuine curiosity, and the mindset that you can “just do things.”

The first one is the most important, and it's verifiable. Exceptional history can show up in sports, school, a professional role, or, in Max's case, competitive gaming. The form doesn’t matter. What does matter is that the person has already proven they push things to the limit.

For curiosity, he asks candidates to tell him about the latest rabbit hole they went down, and to teach him something unusual. A candidate who recently learned to play the violin, for instance, brought it to the interview and walked him through it in real time. Someone who cannot come up with anything “weird,” he finds, tends not to be very curious.

The third quality is a conviction that ambitious things are possible. Max connects this to something he had to unlearn himself. Swedish culture carries a concept called Jantelagen, a kind of social norm against standing out or claiming to be exceptional. At Y Combinator, Legora discovered they had the highest revenue in their batch and were growing the fastest. That external signal helped. But the underlying lesson was that thinking smaller than the opportunity required was a form of waste.

Conclusion

The throughline across Max's thinking is a refusal to optimize locally when a bigger version of the same bet is available. That meant not fine-tuning models when frontier models were improving faster. It meant not selling hard when the product needed more time to be right. It meant treating each new market launch as a repeatable motion rather than a one-off risk. And it meant overcorrecting for ambition once he recognized that thinking small was its own kind of mistake.

Legal AI is still early. The shift from associates to partners using agents is meaningful, but the deeper transformation – legal work happening via agents as often as it happens via humans – is still ahead. Legora's bet is that the firms that trust the platform now will be the ones best positioned when that shift arrives.

The LLM wave, Max says, will only happen once. He intends to ride it.