If you consider the legal tech news of the past few months, even years, you’ll be able to develop a sense of where most big AI projects will head. If it involves BigLaw, it often also involves Harvey or Legora, if it’s eDiscovery, you may well see a Casepoint or Relativity deal, or if it involves personal injury, EvenUp, Eve or Supio could be in the mix. Of course, with new tools and features emerging every day, this is not always the case.
What you do not usually expect is for a mass tort trial lawyer with more than 200,000 clients on the books to review roughly 200 vendors and pick a company whose main line of business isn’t legal tech at all – it’s digital transformation, marketing and technology services for global enterprise brands.
But that is what Mikal Watts, founder of Watts Law Firm LLP and one of the more visible plaintiff-bar trial lawyers in the United States, has done. Since March 2026, Watts has been working with Monks, the operating brand of Sir Martin Sorrell’s S4 Capital Plc, to build the platform he is calling Watts SI. The pair are testing it on the firm’s tariff-refund practice with a public launch imminent, and the longer game is to make the platform available to other plaintiff firms through a management services organisation (or “MSO”).
This week, The Legal Wire sat down with Mikal Watts and Karan Chetal. Karan is the Chief Growth Officer of Monks Technology Services, and the person driving the partnership from the Monks side. We discussed how the two of them ended up in the same room, what they are building, and why.
The 200-vendor test, months of AI school, and a Texan with strong opinions
There is a line in the Bloomberg piece from August 2026 that first put this story on our radar. Mikal, describing his research phase, told the reporter: “There’s so many vendors that are charlatans that will sell you ocean-front property in Arizona.” He is a lawyer who took a four-month course on artificial intelligence at the University of Texas to ensure he came to the table an educated buyer, before hiring anyone to build for him.
Speaking with Karan, he says this shaped how Mikal scoped the project. What Mikal didn’t want was the ability to bolt a chatbot onto an existing platform. What he did want was a fundamental change to the operating ecosystem of his firm.
The scope that Watts articulated narrowed the search fast. “A lot of legal tech tools aren’t purpose-built, but rather commercial off-the-shelf products,” Karan says. “Their proposition is essentially: I’ve built this platform. Bring your business and bring your data, and then operate inside my platform.” A firm choosing that route generally accepts the product as-is. The workflow flexes to fit the software.
What Monks offered was structurally different. Not a product with Watts’s data inside it, but a custom-built platform designed to sit around Watts’s workflow and reshape it end to end. The distinction between the Monks product and other ready-made legal tech tools is what a website would struggle to portray, but it’s essential six months into an implementation, when the product you’ve bought is doing 60% of what you needed it to do and can’t be moved.
With this, Monks also brought a piece of scale credibility most legal tech firms can’t buy. Its existing customer base includes regulated environments where a single event can drive tens of thousands of inbound calls in a day. “Imagine something happens involving a widely used over-the-counter medication and suddenly you’re dealing with 40,000 or 50,000 calls in a single day,” Karan says. “Can your platform handle that kind of scale? Can it operate securely and reliably within that environment?” A plaintiff-bar firm running a mass tort event needs the same shape of infrastructure. Not many pure legal tech vendors have run at that scale in a genuinely regulated environment.
TLW to Mikal: Having reviewed roughly 200 vendors before choosing Monks, what did they miss, and how much of it was an issue with the product vs a willingness to commit to the outcome you wanted?
Mikal: “Of the 200 “AI specialists” I met with, 75% specialized in nothing more than sales gimmickry now clothed as “AI”; 20% more were marketing a solution that latches on to commercially-marketed AI solutions that while excellent, did not provide a closed system that will protect confidential attorney-client information. In short, most Legal-AI products now being sold are using client data to train Large Language Models (LLMs), meaning that attorneys injecting their clients’ data into an off-the-shelf AI solution were per se compromising the sanctity of confidential data they had no right to share with others. This taught me one thing – I needed to custom build a closed platform solution ourselves, not rely on off-the-shelf open platform LLM offerings that would get me in trouble.”
The choice Mikal made is a live one for any firm buying legal AI today. Off-the-shelf products are often cheaper, faster to stand up, and more mature. Custom builds are expensive, take longer from design to implementation, and are specific. Mikal chose expensive and specific for a reason. He was not buying software. He was, in the Monks framing, investing to build an operating model.
Mikal’s decisions will impact how the platform behaves. The Monks build supports the specific workflows of a mass tort, personal injury or tariff recovery practice: qualifying claims (there will always be people trying to game a system, as Karan puts it, so verification matters); triaging documents; drafting litigation with attorney review; and, on the tariff side, an integration with financial institutions so that a claimant who wants to sell their claim for immediate liquidity can get an offer in a single click. None of this is generic.
What a lighter legal tech treatment may miss is that the platform Monks is building for Watts is not meant to remain private. The plan is to use Mikal’s practice as the anchor tenant of an orchestration layer that will be extended to all other plaintiff-bar firms in other practice areas. Immigration is one Karan mentioned as a workflow Mikal himself does not run, but the platform could serve. In his phrasing, they are building the platform, not the product.
The trick, of course, is doing the first build in a way that generalises. A custom platform that only ever serves one firm is an expensive product with a customer count of one. A platform that scales cleanly to 20 is a market.
TLW to Mikal: You took a four-month course on AI at the University of Texas before you signed anything. For plaintiff-bar leaders about to sit through the same wave of vendor pitches you sat through, what does that education teach you to listen for, and what is something that would easily alarm you?
Mikal: “The four-month course was necessary for me to first learn the fundamental building blocks of AI – what is it? How does one construct an AI solution? How does it work? What are the upsides and what are its risks? Just as education, however, was the consistent exposure to weekend courses and YouTube videos on how to use the AI offerings already introduced to me over the past four years – ChatGPT in late-2022, to Harvey AI, Eve, FileVine and CoCounsel in 2023, to Norm AI and Thomson Reuters’ CoCounsel Legal AI in 2024, to Clio, Legora and Disco in 2025, and to the most recent 2026 Legal AI solutions – Chinese Open Source models, Anthropic’s Claude for Legal, OpenAI’s Astra for Law, and Google Gemini’s Enterprise for Legal. The more looked, the more I determined we needed a different approach – Hence, “Watts SI” to be built with Monks.”
The MSO angle, or how private equity found the plaintiff bar
Behind the technology sits an intentional commercial move. Private equity capital has been flowing into the plaintiff bar for well over a year, operating within US restrictions on non-lawyer ownership of law firms through management services organisations. The MSO holds the non-legal functions of a firm (marketing, IT, client intake, back office) while the practice of law stays where it must, with the lawyers.
The investors are getting in on the action. Uplift Investors’ Orion Legal MSO announced its fourth PI firm partner in July 2026. Rafi Law Group announced a $125 million MSO investment in April. Holland & Knight hosted an invite-only conference at its New York office in April with Apollo Global Management, Fortress Investment Group and Stifel Financial Corp all showing interest in the space. The demand side of the market Mikal is aiming at is real, funded, and already assembling.
Mikal’s angle, and Karan’s echo of it, is that these consolidating groups will need a common AI platform sitting across the firms they roll up. And that the firm which supplies the platform gets to define, in a meaningful way, how a large slice of the plaintiff bar operates. It is a big claim, and it puts Monks in the interesting position of not really being in legal tech at all, but in something closer to legal infrastructure.
The commercial appeal for the plaintiff firm is pretty simple. The MSO carries the cost of the AI infrastructure no single small firm could afford alone. “The MSO allows firms to take outside investment, relieving them of the financial burden,” Karan says. That is a genuine unlock for firms that would otherwise be priced out of the sort of build Mikal is funding.
The commercial complication, which nobody in the market seems to have fully addressed yet, is the two-customer problem. When the MSO is buying the platform on behalf of the firms it has invested in, the platform vendor now has two masters. The firm cares about workflow and case outcomes. The MSO cares about returns and scaling economics. Those interests point in the same direction most of the time. Not always.
TLW to Mikal: You have said publicly you are a lawyer and a businessman on the side. Where should the line sit between what an MSO can help a plaintiff firm do faster, and what should stay firmly on the lawyer’s side of the wall?
Mikal: “The fun and important part of being a lawyer is the lawyering, not the managing of ministerial tasks with legions of people serving as a labor force repeating rote steps and tasks that must be completed over the lifespan of a complex case. The concepts of an MSO is allowing others to lift that burden from lawyers who ought to be focused on lawyers. This is where AI can be transformative. I used to employ 55 lawyers and a staff of 260 people; with AI, I now have less than 10 lawyers and a staff of 30 people. AI empowers law firm leaders to focus their efforts on practicing law, not on managing people.”
TLW to Karan: Once the MSO model is fully live, the firm is the practitioner and the PE-backed MSO is the buyer of the tech. Who is the customer for you, and how do you keep the lawyer’s workflow needs and the investor’s return needs from pulling the product in different directions?
Karan: “The law firm is our customer. The MSO may be the one writing the cheque, but a platform that lawyers don’t trust doesn’t generate a return for anyone, so we design from the lawyer’s workflow outward and treat the investor’s needs as outcomes that workflow has to deliver, not as a separate roadmap. In practice, most of the tension goes away once everyone agrees on what’s being measured. The lawyer wants claims qualified accurately and cases moved quickly. The investor wants lower cost per case, faster time to resolution, and the ability to bring the next firm onto the platform without starting from scratch. Those are largely the same metrics viewed from two different chairs. Where they genuinely diverge, the line is clear and it’s built into the architecture. Anything that is a legal judgment, whether a case is taken, how it’s pleaded, when it settles, stays with the attorney, and every action on a matter is traceable to the person or agent that took it. The MSO gets shared infrastructure, a common orchestration layer and visibility into performance across its portfolio. It doesn’t get a say in how law is practiced.”
Outcomes over widgets: the contract most vendors won’t sign
Karan makes some really useful observations regarding how Monks contracts. Most enterprise software is sold on availability: uptime guarantees, feature guarantees, service-level agreements around response times. Monks is signing what Karan calls outcome-based contracts, structured around the business result the firm is trying to achieve. How quickly the claim gets recovered. How quickly the claimant sees the money. How quickly a case gets qualified.
“We don’t sign what I would call hope-based contracts. We sign outcome-based contracts. We deliver to the business outcome.” That is a meaningfully different commercial posture, and it changes what the buyer is paying for. The vendor does not get to ship the software, invoice, and disappear.
TLW to Karan: What does ‘missing the outcome’ look like commercially on Monks’ side, and how has pricing that risk changed the way you scope a legal deal?
Karan: “Commercially, it means a meaningful share of our fees is at risk against the results we’ve committed to. If the outcome doesn’t land, Monks absorbs that, not the client. That’s the point. A licence vendor gets paid whether or not the firm ever sees the value. Pricing that risk has made us far more disciplined before we sign. We spend real time in discovery mapping the actual workflow, the quality of the data, where the bottlenecks are and which of them are genuinely solvable. We then agree on a small number of measurable outcomes, such as time to qualify a claim, time to recovery and cost per case, and we baseline them before we build. We also phase the work, so the first release proves the outcome in one practice area, as we’re doing with tariff recovery, before we extend it. And it makes us selective. If a firm wants a tool but isn’t willing to change the process around it, we can’t credibly underwrite an outcome, and we’d rather say that early than sign a hope-based contract.”
The legal tech blind spot and traceability
Two smaller observations from Karan really caught my attention. Both reflect where Monks’ non-legal background shows up plainly in the platform.
The first is audit-trail depth. Monks built the platform so that every action on a matter is traceable to the entity that took it: lawyer, paralegal, or AI agent. That capability is more common in regulated industries than in legal tech, where a great deal of AI shipping today has audit-trail depth that wouldn’t pass muster with a financial-services CIO. For a plaintiff firm that will one day need to defend how a case was handled, this is not a nice-to-have.
The second is the customer-experience layer. Karan critiques the traditional plaintiff-bar model, which many vendors avoid doing in public. Historically, he says, law firms have not had the tools to measure client satisfaction end to end. The Monks build gives Watts a way to measure client satisfaction end to end. That will not sound like a revolution to anyone who has worked in a consumer business. It is meaningfully new inside a mass tort firm.
Neither of these features would sit near the top of a legal-tech-first vendor’s roadmap. They are here because the people building the platform come from an industry where both were table stakes.
Where this fits
The Watts and Monks partnership is an interesting development in plaintiff-bar AI, which in and of itself is receiving substantial attention right now. The vendor is not a legal tech company. The customer is not BigLaw. The commercial structure is not a licence deal. The exit path is a platform a still-forming MSO market will eventually route through.
That combination is something we’ll be keeping an eye on.
Firstly, because of what it says about who the plaintiff bar is going to buy from over the next two years. Mikal is a sophisticated buyer in that market, and he picked a vendor most (if not all) of the market would not have had on a shortlist. If the bet pays off, that decision will be studied.
Secondly, because of what it says about where the durable value in legal AI can be found. Monks is not competing on model capability, and does not seem to be trying to. What they are selling is workflow, orchestration, and a contract structure that promises the outcome. If the past year of legal tech commentary was mostly about models and platforms, the next may be more about who is willing to sign for the result.
Whether Watts SI itself scales into the MSO market is a separate question. The pattern the partnership is testing, however, is one the wider market has not yet fully absorbed. Outcome contracts. Custom-built orchestration. A non-legal-tech vendor entering through the plaintiff bar’s side door.
