Why We Invested in Decover AI?
For the longest time, legal tech had an unglamorous reputation among VCs. Its core market — conservative, highly regulated law firms — was viewed as resistant to change, casting doubt on whether new tools could gain broad adoption. Yet, in the past couple of years, this dynamic has shifted. We’ve seen a wave of legal tech companies reaching significant commercial milestones, especially those using LLMs in targeted, high-impact workflows. In a sea of AI startups still searching for PMF, legal tech has emerged as a rare oasis of tangible demand.
Today, we’re excited to announce our investment in Decover AI — a company that started by reimagining e-discovery, one of the most time and cost-intensive corners of the legal world, and is building a broader platform for “legal intelligence.” After speaking with countless lawyers, operators, and in-house counsel, the consensus is clear: AI is no longer optional; it’s a necessity. Decover AI sits squarely in the middle of this transition, offering solutions that not only cut e-discovery costs dramatically but also pave the way for more advanced research and knowledge management capabilities.
E-Discovery as a wedge
The entry point for Decover AI is e-discovery — a process that compels lawyers to sift through mountains of digital files (emails, chats, audio transcripts, and social media) in search of a potential “smoking gun.” With the global datasphere projected to reach 175 zettabytes by 2025, even routine civil cases can involve hundreds of gigabytes of data. Existing tools like Relativity, Nextpoint, CS Disco, and Logikcull were instrumental in digitising e-discovery but often rely on rigid keyword searches. The results can be unwieldy: too many irrelevant hits, or missed documents hidden by subtle language cues.
For large firms, running these searches in-house can be so cumbersome that they outsource to specialised legal service providers (LSPs). By the time a matter concludes, the bill for document review alone can climb into the seven or eight-figure range. Decover AI tackles these pain points head-on with a hybrid approach that blends classic lexical methods (to ensure accuracy on critical terminology) with advanced semantic search (to capture meaning and context). Rather than forcing users to build complex queries, Decover AI’s natural language interface allows lawyers to simply ask, “Show me all emails connecting Person A with the marketing department’s budget changes in the last six months.”
In pilot tests, partners have reported that e-discovery tasks, which once dragged on for days, can be compressed into minutes. Smaller law firms see this as a chance to level the playing field, while larger ones are eager to streamline the process in-house, saving on outside vendor fees.
Beyond the Search: A Broader Legal “Brain”
If Decover AI were merely a best-in-class e-discovery engine, that would be impressive enough. But their roadmap extends further. The founders aim to unify e-discovery, legal research, and knowledge management into an integrated, AI-driven “legal intelligence” platform.
Why does this matter? Anyone who has practised law knows it’s not just about finding documents; it’s about constructing compelling arguments. That entails referencing statutes, synthesising case law, and pulling insights from internal repositories of motions, briefs, and firm-specific wisdom — assets scattered across countless tools.
Decover AI’s “generative defence” feature offers a glimpse of what unified legal intelligence could look like. Imagine asking, “What’s our best strategy for a commercial lease dispute under California law, given how we’ve argued similar cases before Judge X?” Rather than just presenting a file dump, Decover AI synthesises key documents from e-discovery, taps external databases for relevant case law, and incorporates the firm’s own litigation history. The output isn’t just a raw data dump — it’s a thoughtful draft that reads more like a senior attorney’s annotated notes.
A Founding Team Built for the Job
Conquering the legal sector’s complexities demands more than a sophisticated product; it requires a keen sense of enterprise sales cycles, trust-building, and stringent data compliance. Decover AI’s leadership team — Ravi Tandon, Kevin J. Van Horn, and Janar Ramalingam — brings precisely that combination of technical and commercial expertise.
- Ravi Tandon spent nearly a decade at ThoughtSpot, honing enterprise-scale search solutions. He witnessed firsthand how a powerful search engine can transform corporate data analysis and boost productivity.
- Kevin J. Van Horn has a diverse background spanning lobbying on Capitol Hill, hedge fund operations, and driving AI-focused initiatives in defence and intelligence. That experience in navigating complex deals and procurement processes will be critical for selling to large law firms and federal institutions.
- Janar Ramalingam spent over 12 years at Yahoo, building large-scale systems and working on advanced machine-learning algorithms. He later joined ThoughtSpot as a principal engineer, refining his data-centric product skills alongside Ravi.
While many AI startups struggle to build a moat, verticalised AI companies can develop defensible advantages through deep domain knowledge and strong buyer insights. In legal tech, the chief competitor is often not another platform; it’s inertia — law firms still tethered to clunky legacy tools or even spreadsheets. Decover’s founders understand how attorneys and their clients buy software, and they have the grit and credibility needed to win over a market where trust is paramount.
Why Now?
Legal tech has historically faced an uphill battle. Institutional inertia, regulatory intricacies, and the entrenched billable-hour model made efficiency gains less urgent. Yet, in the past two years, we have seen a new wave of AI solutions breakthroughs, establishing genuine PMF where other LLM-based ventures have fallen short. We see four forces behind this shift:
- Lawyers work with enormous volumes of unstructured text — depositions, briefs, correspondence — and LLMs excel at processing exactly that type of data.
- Law firms are feeling competitive pressure to contain costs and boost throughput, especially smaller practices aiming to stand out in a crowded market.
- High-profile deals (like Harvey’s notable partnership with Allen & Overy) have validated AI-based solutions in law, prompting even traditional holdouts to experiment.
- Data volumes keep climbing: e-discovery can easily run into hundreds of GBs of documents today on average, and that figure is set only to multiply. Lawyers can’t keep up without AI-driven automation.
Decover AI’s Early Impact and Industry Focus
Even in its early stages, Decover AI has processed over 30M documents, delivering an estimated $4.5M in direct cost savings for legal teams and unlocking at least $15M in expedited settlements. This impact spans multiple phases of litigation. In medical malpractice cases, for example, building automated medical chronologies once took months — Decover AI can do it in minutes, increasing the volume of cases that firms can handle by about 10%. Large-scale white-collar crime matters, often involving millions of records, have seen an 80% reduction in document review time thanks to automated e-discovery. In legal research, the platform has cut the time to find key precedents by a factor of five, drastically reducing the risk of missing critical references.
One particularly promising area is construction defect litigation, with specialised documents, numerous subcontractors, and deep-pocketed insurers — all factors that inflate the cost of manual review. Insurers routinely pay premium rates for fact-finding in these cases. A robust AI platform that slices through large data sets quickly delivers a significant ROI, turning Decover into a true strategic advantage rather than a mere convenience.
It is also worth noting that the adoption model for e-discovery is predominantly case-by-case. Large law firms frequently deploy Relativity for one matter and Nextpoint for another, rather than committing a firm-wide mandate to any single provider. This piecemeal approach lowers the barriers for a new entrant; they don’t need to displace an incumbent across every practice group, only demonstrate clear value on a specific case. Once that first case proves successful — fast searches, minimal false positives — lawyers are naturally inclined to bring Decover AI into subsequent matters.
Looking Ahead
The U.S. legal tech market is predicted to surpass $15 billion by 2026, with e-discovery alone accounting for roughly $4 billion. However, the broader opportunity is even greater when you consider the trillion-dollar global legal services market. If Decover AI proves indispensable in one practice area, it can seamlessly extend into research, drafting, knowledge management, and other workflows — capturing a far larger market than any single-point solution.
We’re thrilled to support the Decover AI team on this journey. Their blend of technical excellence, market savvy, and genuine passion for reinventing how lawyers work places them in a prime position to shape the future of legal practice. We believe Decover AI will help usher in a new era of value and efficiency for law firms and clients alike — an evolution that’s been overdue for decades.
