Robert Ambrogi Blog
At the recent Legalweek conference in New York, I sat down with the cofounders of Syllo, a legal technology company that has taken a uniquely ambitious approach to developing technology for litigators.
Rather than focus on any one segment of the litigation process, they have built a unified, AI-powered litigation workspace that spans the entire lifecycle of a case, from raw data collection through trial.
The company is the brainchild of Jeffrey Chivers, its chief executive officer and head of product, and Theodore Rostow, its chief operating officer and chief strategy officer. The pair cofounded the company in 2019 after meeting while clerking in federal court.
Chivers, a Harvard computer science graduate who went on to earn his law degree at Georgetown and litigate at Sullivan & Cromwell, says he brings a technologist’s instincts to the practice of law.
Rostow, a Yale Law School graduate who practiced at Simpson Thacher & Bartlett, had been immersed in legal technology and AI since 2014, writing about automation, structured and unstructured data, and deep learning while still in law school.
The Monastic Period
Part of what makes their story unusual – and what may make it instructive for anyone watching the trajectory of legal technology – is how they chose to build the company.
Rather than follow the conventional startup playbook of building a narrow point solution and expanding from there, Chivers and Rostow decided from the outset to build the entire infrastructure for a unified litigation platform.
That meant years of foundational work on plumbing and architecture before any meaningful user interface existed on top of it.
“We basically decided we are going to build the unified platform,” Chivers told me. “And we didn’t build one thing or another thing and then add other things on. We actually built out the entire infrastructure of the entire unified platform, which did almost nothing.”
That approach came at a cost. No venture capital firm wanted to fund a multi-year R&D project with no near-term product to show.
To keep the lights on, Chivers and Rostow ran a litigation practice alongside the company for more than three years. Their initial investors were lawyers and high-net-worth individuals who understood the long-term vision – people who, as Chivers put it, “knew a ton about the legal industry and could really understand the long-term value proposition.”
The company’s early philosophy, Rostow said, was simple: “We want the risk to be placed entirely on, ‘Can we actually build this thing?’ If we build it, it’s definitely the right thing.”
The technical foundation was built in collaboration with engineers from Carnegie Mellon University, including Jamie Callan, professor at CMU’s Language Technologies Institute, who collaborated with the company on its agentic AI document review research.
What Syllo Actually Does
Today, Syllo describes itself as the broadest litigation technology tool on the market. Based on the capabilities that Chivers and Rostow walked me through, but which I have not yet seen demonstrated, that claim is at least plausible.
The platform encompasses:
- E-discovery, including raw data ingestion, load file ingestion, search, threading, deduplication, redaction, and production.
- Agentic AI document review, which automates responsiveness review, issue coding, privilege review, confidentiality review, and PII/PHI review and redaction. The system can apply an unlimited number of issue codes across data sets of any size.
- Case building and fact management, including exhibit management, transcript management, pretrial document stamping, and deposition preparation.
- A litigation agent that can draw on everything across the platform for dialogue-based work, analysis, and work-product generation.
That combination of features means that Syllo covers not just the EDRM (Electronic Discovery Reference Model), but extends beyond it into case strategy, trial preparation and document generation.
The company competes, depending on the use case, with e-discovery platforms such as Relativity, Everlaw, DISCO and Reveal; with e-discovery services companies such as Consilio, Epiq and FTI; with case-building tools such as Opus 2; and with AI legal assistants such as Harvey and CoCounsel.
Chivers acknowledged that some firms have trouble grasping that broad competitive positioning. “You could imagine we go in and tell a firm, use Syllo instead of using Relativity, Opus 2 and CoCounsel or Harvey. That is really hard for a law firm to accept.”
Why the Unified Approach?
By no means would Syllo be the first legal tech company to attempt to unify the litigation stack. Both LexisNexis and Thomson Reuters have tried it over the years through various acquisitions and product combinations, and many e-discovery providers today are pushing their products farther to the right of the EDRM into case management and trial preparation.
But Chivers sees those attempts as largely having failed, and he attributes those failures partly to the different kinds of software engineering required by each domain.
E-discovery involves forensically sound data processing, complex metadata handling and constantly evolving technical standards. Case management involves more lawyer-facing workflow and strategic tools. Building both under one roof requires distinct engineering competencies and a single data environment that connects them.
It is precisely that single data environment, Chivers argued, that makes AI maximally effective.
“What I think people are starting to appreciate more,” he said, “is that to get the most out of AI and agentic AI involves structured data. It involves extremely clean data. It involves very elaborate context management by the system itself.”
The way Syllo was built, he added, “was exactly what is needed to get the most out of AI.”
The practical implications are significant, he said. Because Syllo houses everything from the raw collected data through to case strategy, its AI agent can draw on the entire factual record when generating analysis or work product, not just the subset of data a lawyer has manually fed into a standalone AI tool.
Agentic Document Review
While Syllo was built as a full-lifecycle platform, the feature that it first brought to market was its agentic AI document review system, which the company began deploying with clients in February 2024.
The system coordinates multiple large language models that autonomously manage and execute document review tasks based on user-defined guidelines, the company says. It dynamically allocates more powerful models to complex documents and simpler models to straightforward ones, and it can handle an unlimited number of issue codes without the prompt-overload problems that plague more linear applications of AI to large data sets.
In March 2025, Syllo and a group of 25 litigators and e-discovery practitioners from seven law firms published a white paper, in collaboration with Carnegie Mellon’s Callan, documenting how the system performs in active litigation.
The paper reported that the system routinely achieves estimated recall above 93 percent, with an average recall of 97.8%. The median estimated precision was 85.9%, and the average estimated precision was 79.7%.
In one example discussed in the white paper, a Quinn Emanuel trial team used Syllo in a merger enforcement case where the firm was retained just weeks before trial. The team used Syllo for outgoing and incoming document review, deposition preparation and case strategy across more than 70,000 documents. Quinn Emanuel won the trial, and the team was named AmLaw Litigators of the Week in March 2025.
Chivers told me the agentic document review capability was what initially drew many AmLaw 100 firms to the platform. “Firms will come in and think Syllo is good at document review, and that’s why they want to work with us,” he said. “And then they’ll onboard the case and realize, ‘Oh, Syllo is an entire platform for all this stuff.’”
Clients and Market Position
Syllo’s customer base skews heavily toward the larger-firm market. The company counts multiple Am Law 100 firms among its users, including Quinn Emanuel Urquhart & Sullivan, Pillsbury Winthrop Shaw Pittman (which announced a multi-year integration of the platform last year), and Ballard Spahr.
Its sweet spot, Chivers and Rostow said, is complex, high-stakes litigation – think antitrust MDLs, environmental litigation, complex bankruptcy, patent cases – where the sheer volume and complexity of the factual record can be overwhelming.
Syllo has also been expanding beyond Big Law. In 2024, it announced a partnership with the National Employment Lawyers Association (NELA) to make its platform available to the organization’s 4,000-plus members, many of whom are plaintiffs’ employment lawyers at smaller firms.
From Stealth to Scale
For much of its existence, Syllo operated in near-total stealth, bound by NDAs with its law firm clients. The company had no salespeople for its first five years and, as recently as late last year, employed only three.
That is changing. Last October, the company announced a $30 million growth funding round led by Venrock, with participation from Two Seas Capital and strategic investors. Venrock partner Nick Beim said the company’s software represents “a step-function improvement in capabilities for the industry.”
Since the funding round, the company has been growing rapidly. Chivers told me that Syllo has roughly doubled its headcount in just the past four months, adding approximately 30 people, and was about to bring on a head of sales.
In February, the company announced the hiring of Alexis Mitchell as senior AI consultant. Mitchell previously held key data science and linguistic roles at Epiq, Opus 2, OpenText, Recommind, Autonomy and H5. That hire came alongside a broader talent expansion drawing from companies such as Meta, SpaceX, Amazon, Sony, Google and OpenAI.
“On the product side, we are now at a decisive lead moment,” Rostow told me. “The question is how, over the next year and a half, does that lead dramatically widen, even as hundreds of millions of dollars and dozens of billions of dollars are poured into the space.”
Whether Syllo can translate its technical breadth into sustained commercial growth in a market that traditionally rewards specialization remains to be seen.
But after five years of building in stealth, a successful funding round, and a growing roster of elite law firm clients, they appear to be well-positioned to give it a shot.