The Terms Were the Terms
Richard Perris spent 15 years as General Counsel of CVC Capital Partners—on the side of private markets deals that held all the power. What he saw from that seat, and what he decided to do about it, is what led to Covenant.
Richard Perris spent 15 years watching private markets deals get done from one of the most powerful seats in the room. As General Counsel of CVC Capital Partners, he was on the side that held all the cards. Funds were oversubscribed. Investors lined up. The terms were the terms.
It was only when he left that the other side of that picture came fully into focus.
"We'd send them 300 pages of complicated legal documents and say, read it yourself," he says. "Oh, by the way, you're paying the bill for our lawyers as well." The investors on the receiving end faced a stark choice: hire their own specialist counsel, which would be expensive, slow, and unlikely to change anything, or close their eyes and sign.
The economics of the situation made it structurally difficult for investors to push back. "If investors wanted to even understand the terms, they needed to hire their own attorneys just to read and explain them," he says. The negotiating dynamic between large fund managers and their investors, shaped by decades of market convention and significant imbalances in scale and resources, meant that terms often went unchallenged largely because the cost and effort of challenging them was prohibitive.
That frustration and the structural conditions that created it are what Richard spent the years after CVC trying to solve. And it's what eventually became the foundation of Covenant.
A problem worth building for
Richard hadn't planned to build a company. After leaving CVC, he spent time exploring, including a brief foray into crypto. But he'd been watching AI develop since the early days, taking demos from startups trying to review NDAs with models that, as he puts it, "couldn't tell the difference between a cat and a dog." None of it worked. But he kept taking the meetings.
By the time GPT-3 arrived, something had shifted. The technology was starting to look like it might actually work—like it could understand complicated things, not just pattern-match on simple ones. And Richard had a very specific problem in mind.
"Combining the two ideas: you've got this technology that does seem to be able to understand things that are quite complicated. Could you use that to solve the problem I'd identified?"
The problem was the one he'd watched play out from the other side of the table: investors in private markets deals who needed smart, fast, affordable legal advice on highly specialized documents, and had no good way to get it.
He started talking to everyone he knew. Most thought it was a reasonable idea. But it wasn't until he was introduced to Jen Berrent who was building a law firm around AI with a very specific go-to-market strategy that it clicked.
What the technology actually changes
Five years ago, reviewing a fund document meant printing it out and starting on page one. "I can't imagine doing that now," Richard says. "It would take ten hours." Today, Covenant runs documents through a series of AI models that do the initial review—flagging issues, identifying key terms, surfacing what matters—before a lawyer touches it. "It's like having this amazing team of associates who've gone through it in quite a lot of detail, way more detail than any human would bother to do. You get this huge leg up."
One of Covenant's clients is a medium-sized endowment with a lean team that had long wanted to run a co-investment program. Co-investments—direct stakes in deals alongside a GP, often fee-free—are among the most attractive opportunities available to private market investors. But they come with a catch: GPs need an answer fast. The deals move quickly, the documents are complex, and any investor who can't commit and execute promptly simply gets passed over.
“An ability to review the documents and flag any issues quickly is paramount," says Richard. With traditional outside counsel, the costs were high and unpredictable—exactly the wrong profile for a budget-conscious endowment trying to build a new program. "Our client needed external advisers to help them execute the deals but did not have unlimited budget. Our system gives them full predictability on both turnaround time and cost."
The human element
There's a question that comes up constantly in conversations about AI and legal services: where do human lawyers fit? "They want to know there's a human being who is saying this is correct," Richard says. "The AI becomes part of the team." The work gets done faster and more thoroughly. The judgment and accountability remain human. For investors who have long needed better access to specialized legal support, that combination is the point.
The market is still catching up
When Richard talks to prospects, the objections are rarely about whether AI can do the work. Clients—the investors writing checks into private markets funds—have watched legal costs compound for years. They don't need convincing that much of that work is automatable.
The skepticism runs elsewhere. Data security. Model transparency. The organizational question of whether this is a decision for the general counsel or the CTO. And the fundamental confusion about what Covenant actually is—not quite a software product, not quite a traditional law firm, something the market is still learning how to categorize.
"No one really knows what things are going to look like in five years," he says. "But the questions we're getting are: Is this really going to benefit me? What am I actually buying? Am I buying software? Am I just switching my law firm to someone else?"
It's a fair set of questions for a category that doesn't yet have a name. Richard is patient about it. He spent years on the side of the table that didn't need to explain itself. Now he's building for the side that was always waiting for someone to.
