Legaltech Aavalynx has raised £1.5 million in pre-seed funding to help enterprises tackle the financial risk of legal disputes.
Founded in 2023 and commercially live in 2024, Aavalynx has developed proprietary AI tech that enables legal teams, law firms and litigation funders to structure, analyse, and interrogate dispute data at scale. It helps companies intervene earlier, make better-informed strategic decisions, improve litigation outcomes, and regain control over legal spend.
I spoke to Hanna Roos, founder and CEO of Aavalynx, to learn more.
The hidden cost of corporate litigation
Today, about 30 per cent of total corporate external legal spend is on litigation, and 80 per cent of companies were involved in at least one lawsuit in 2025, with more than 46 per cent expecting an increase in lawsuits, according to a 2026 study survey of US corporate counsel.
Despite the scale of this risk, organisations still lack the structured data and tools to systematically analyse dispute portfolios, identify emerging trends, or forecast likely outcomes with accuracy. As a result, they remain heavily dependent on external law firms for strategic decision-making.
Yet classic law firm models are not equipped to provide a holistic risk assessment, for example, the impact on reputation and customer trust.
Roos brings almost two decades of legal experience, having led complex, high-stakes international disputes in roles at Freshfields, Latham & Watkins, and Quinn Emanuel.
Roos says the idea for Aavalynx emerged after years of seeing the same structural problems repeated across high-value disputes.
“I loved the business of disputes. At university I was fascinated by Plato and Aristotle and their vision of how to live well, and by the idea that you could help deliver justice internationally through global disputes.”
But once she started practising, she realised there were three fundamental problems. The first concerns law firms and the difference between the business of law, which is what law firms do, and law for business, which is what in-house legal teams need to deliver.
“Traditional law firm models are based on volume and hourly billing. Clients don’t pass resolutions based on volume, but if your business model depends on volume, that’s what happens.
The average dispute involving states lasts 4.8 years. That’s an average, and it’s incredibly high.”
Disputes often run for several years, if not decades, and can cost enterprise clients hundreds of millions per year.
The total exposure of ongoing cases can greatly exceed a company’s annual revenue. In some cases, law firm profit margins can approach 80 per cent.
Further, according to Roos, the larger the company’s revenue, the larger its disputes portfolio.
“To anonymise actual figures, you might have a business generating $50 billion in revenue with annual legal spending on disputes alone of around $200 million.”
She further shared that if you compare the size of a company’s disputes portfolio, particularly in sectors like defence, that portfolio is effectively a risk asset.
“The total exposure can greatly exceed annual turnover, meaning the company is carrying a significant sub-optimal risk asset on its balance sheet.”
The second problem is with in-house legal teams. She contends that when LLMs first emerged, her team started looking at the in-house space and realised many organisations didn’t readily know where their data sat or how to harness it effectively.
“That leads to a lack of transparency, slower decision-making and sub-optimal processes.”
The third issue is the lawyers themselves. Roos contends that law is a sector built entirely around talent, but it doesn’t always recognise that.
“If you harness AI properly, combine it with good leadership and bring cognitive science into the way teams operate, suddenly you can deliver brilliant results without humans burning out.
My tagline is “bringing joy back to lawyers.””
Building AI to prevent disputes, not just manage them
Currently, there’s an abundance of legal AI tools from companies such as Harvey and Legora designed to improve lawyer productivity. Roos believes Aavalynx is solving a different problem.
“Current AI tools by and large improve lawyer efficiency,” said Roos.
“But efficiency is not good enough. Generic LLM models cannot address the unique complexities of a sector like dispute resolution.
Good tools make disputes efficient, but great ones make them disappear. Even affirmative litigation (for a tech company pushing licence fees) should be as short and impactful as possible.”
Aavalynx gives companies the data and visibility to make better decisions earlier in the lifecycle of a dispute, where the financial impact is greatest. That shifts them from reactive firefighting to proactive decision-making. If clients can better understand their data, dispute length and cost won’t run wild.”
Today, organisations have entire ecosystems involved in disputes, including external law firms, in-house teams, financial experts and local legal teams. “It’s not unusual to have twenty people working on a single matter,” Roos reveals. Avalynx aims to provide one central repository for information on a dispute.
“I think of Aavalynx as the control panel of an aircraft. If you’re the pilot, what information do you need in front of you to fly safely? We bring everyone onto that platform so they can make more strategic decisions and resolve disputes faster.”
Making the business case for legal AI
When it comes to measuring the success of legal AI, Roos argues there are two distinct levels.
“The first is straightforward: if you’re the Head of Disputes or General Counsel, your role is to solve difficult business problems. The question is whether you’re resolving disputes more effectively and what financial impact that creates.
The second is demonstrating the overall value the legal function delivers to the business.”
Roos believes this is particularly important for in-house legal teams, which are often viewed as a cost centre despite creating significant commercial value.
“The best lawyers in private practice are seen as revenue generators. When those same brilliant lawyers move in-house, they suddenly become part of the back office and are viewed as a cost. But they’re creating enormous value for the organisation.”
Legal AI can help quantify that contribution through measurable outcomes.
“If you can resolve disputes three years earlier than expected, reduce legal fees by $30 million, save another $40 million on quantum, or preserve a contract by avoiding reputational damage, you’ve demonstrated value not just on individual disputes but across the entire portfolio.”
The data allows a team to identify patterns:
“Perhaps procurement is responsible for a disproportionate number of disputes. You can work with that team, improve the underlying business process and prevent those disputes from arising in the future.
Ultimately, it’s about putting in-house legal teams in a position where they can articulate the value they create using data.”
Roos further says enterprise clients increasingly want AI investments to deliver measurable commercial outcomes.
“Clients want to know: what savings have been generated for us? How has AI improved outcomes?”
As a result, Aavalynx is developing a billing analytics capability. Clients will be able to understand how work is being delivered and where value is being created, rather than simply looking at hours billed.
Legal AI needs evidence, not hallucinations
Aavalynx was designed to minimise hallucinations by grounding its AI in evidence rather than relying on the model’s general knowledge. Roos recalled her own experience, sharing:
“I was working on a live dispute when a client called and said, ‘I’ve proved that the chair of this tribunal is corrupt. ChatGPT found two articles they co-authored with the opposing lawyer’s spouse.”
The articles looked completely legitimate. They included citations and references to what appeared to be leading arbitration sources. But she couldn’t find them anywhere and, after an hour and a half of fact-checking with contacts in the field, concluded that the articles simply did not exist.
Instead of relying on a model’s general knowledge, Aavalynx analyses only the relevant dispute documents supplied by the user before generating an answer. Every statement is linked back to its source through detailed citations, allowing lawyers to open the document and jump directly to the supporting passage.
Roos, however, contends that lawyers must avoid “cognitive surrender” by remaining actively involved in reviewing and validating the AI’s work. “Lawyers have to stay alert, remain in the loop and verify the outputs. That is absolutely crucial.”
The round was led by European Omega Ventures, with participation from West Coast-based Two Ravens and prominent angel investors, including senior law firm partners and a former Amazon European head.
“We’ve seen a new generation of AI platforms attract significant capital over the past 12 months”, said Birgir Ragnarsson, a Founding Partner at Omega Ventures.
“There is a lot of noise in the legal AI space, and much of this activity has focused on improving productivity within legal teams.
Aavalynx is built around a more impactful premise: that the primary challenge for large organisations is not efficiency alone, but the ability to understand and manage disputes as a financial risk. As the legal AI market matures, the winners will be the solutions that can demonstrate clear commercial impact.”
Isaac Kato, a Founding Partner at Two Ravens, adds:
“Hanna is an extraordinary founder with deep disputes expertise who has amalgamated a team of internationally ranked disputes lawyers and leading Nordic tech specialists. It’s clear to us that Aavalynx is best positioned to lead that next phase of disruption.”
Choosing investors for a global company
In terms of fundraising experience, Roos admits it was a “completely new world.”
“There was a huge amount to learn. I realised fundraising is a galaxy of its own, with its own jargon, mindset and priorities. One thing I learnt quite quickly was to identify the people I wanted to work with.”
Roos says she looked for three qualities in investors. First, they needed to be “fiercely intelligent”. Further, they needed to bring genuine specialism — whether that was domain expertise, client expertise or technology scaling expertise.
“And they needed to be kind. I spent a lot of time on that journey and had some great conversations. I’m incredibly happy with the investor group we’ve built for the pre-seed round.”
It was important to Aavalynx to raise from an international investor base because the company is building a global business. Roos explained:
“Our clients are global. Their disputes are global. The underlying issues are remarkably similar whether you’re dealing with a dispute in Japan or California. We’re always looking for global talent and globally experienced investors.”
Momentum has been strong since Aavalynx launched in 2023, with co-development partners including Vodafone. Early data shows that Aavalynx delivers an ROI of ca 30x in saved damages, legal fees, and interest. The figure is as high as 200x when including the rescued commercial opportunities.
Aavalynx will use the funding to accelerate product development, expand the team and fast-track the company’s mission to bring “death to disputes”.