Founder / Legal AI
Sara Dwyer
She taught software to read a medical chart the way a trial lawyer would - and a six-million-dollar seed round followed.
Sara Dwyer, co-founder & CEO, Parambil - New York
She taught software to read a medical chart the way a trial lawyer would - and a six-million-dollar seed round followed.
Those three questions sit at the center of every injury case in America, and for decades the answers were buried in paper. Sara Dwyer built a company to dig them out. Parambil, the New York startup she co-founded and runs, points artificial intelligence at the messiest documents in the legal system - hospital charts, pharmacy logs, deposition transcripts - and hands a law firm something close to a verdict-ready understanding of its own case.
The pitch is deceptively plain. A single mass tort or malpractice matter can run past 60,000 pages. Somewhere in that stack is the missed diagnosis, the undocumented deviation from the standard of care, the detail that moves a settlement by millions of dollars. Humans miss it because humans get tired. Parambil's software does not. It reads everything, cites everything, and flags the moments a partner needs to see.
That is the part most people get wrong about her work. It is not a faster photocopier. It is a reasoning engine for litigation, trained to behave less like a search box and more like a very patient associate who has read the entire file twice before the meeting starts.
Before Parambil, Dwyer spent nearly four years at McKinsey & Company as an engagement manager working where artificial intelligence meets life sciences. It is the kind of seat that teaches you to recognize a real problem from a fashionable one. The real problem she kept circling was this: the law was drowning in medical detail, and nobody had built a tool that actually understood both languages at once.
Her own training made her unusually suited to the bilingual job. She studied computer science at the University of Pennsylvania - and, in a flourish that says something about her appetite for difficult things, minored in Mandarin. She left consulting in early 2023 to start the company.
She did not do it alone, and she likes to say so. Parambil's founding bench reads like a deliberate collision of worlds: a founding engineer steeped in clinical research and big data, and Dr. Ralph Horwitz, a physician who spent three decades in medicine and led the department at both Yale and Stanford. McKinsey, Stanford Medicine, Yale Medicine - the resumes were assembled to cover the exact seams where legal and medical reasoning fail to meet.
It is a telling way to describe a litigation company. Not a weapon, not an edge - a tool for resolution. Parambil works for both sides of the aisle, plaintiff and defense, because the underlying job never changes regardless of who is asking. Understand the medicine. Map it to the law. Price the truth.
In late 2025 the company moved from clever software to something more ambitious: an agentic platform, a set of specialized AI agents that do the work without waiting to be asked. Each has a name and a job.
Runs end-to-end case investigation on its own - searching records, cross-referencing clinical guidelines, flagging gaps in documentation, all with citations.
Turns verified case data into grounded first drafts of complaints, demands and discovery requests. Hours of legal writing become minutes, exported as editable Word files.
Keeps tens of thousands of documents in order automatically - grouped by defendant, facility, or any framework the attorney defines.
The distinction Dwyer draws, again and again, is about trust. This is not probabilistic AI optimized for speed and willing to guess. It is built for defensible accuracy - the kind a partner can stand behind in front of a judge. In a field where a single confident hallucination can sink a filing, that restraint is the whole product.
The results her firm reports back are the sort that make investors lean forward: intake-to-evaluation timelines compressed from months to under two weeks, case review time cut by more than ninety percent, and - the line that matters most - high-value insights surfaced that can separate a seven-figure verdict from a fraction of it.
The funding arrived in a tidy one-two. In January 2025, Daybreak Ventures led a $2M pre-seed. Daybreak's Rex Woodbury, announcing the deal, called Dwyer a force of nature - the kind of unguarded praise investors usually save for their winners.
Exactly a year later, in January 2026, Bling Capital led a $6M seed, with NVP Capital and a roster of angels who actually live inside mass torts, personal injury and legal tech. The money is pointed at the obvious places: more engineering, deeper domain-specific workflows, and the customer-success muscle needed to keep hundreds of demanding law firms happy at once.
What the rounds really bought was conviction. Parambil's wager is that the most human corner of the law - the place where a missed detail changes a life - is exactly where careful AI belongs. And that improving how litigation reads medicine quietly improves healthcare too, by making its patterns legible.
Profile built from public sources: Parambil, LinkedIn, Crunchbase, The Org, PR Newswire, Law.com, The SaaS News. Facts current as of June 2026.