Founding Legal Engineer | Epiq | USA
Anna Teresita (“Teri”) Marcelo is the Founding Legal Engineer at Epiq AI Labs, Epiq’s
advanced research and development group. She works with lawyers, scientists,
engineers, and product teams to translate legal workflows into AI systems and evaluate
those systems for accuracy, reliability, and defensibility. Her work spans litigation,
investigations, and corporate legal matters, and includes direct collaboration with in-
house legal teams and law firms.
Previously, Teri was a Legal Engineer at Norm Ai, where she translated regulatory
requirements into machine-readable logic for AI-assisted compliance and helped lead
enterprise deployments. Before moving into Legal AI, Teri practiced complex civil and commercial litigation in
New York, representing clients in state and federal courts and in arbitration. She earlier
clerked for two Justices of the Supreme Court of the Philippines, drafting decisions
involving anti-corruption, arbitration, and constitutional law and supporting the
enforcement of the South China Sea Arbitration. She began her career at SyCip Salazar
Hernandez & Gatmaitan.
Teri holds an LL.M. from Harvard Law School and a J.D. from Ateneo de Manila
University School of Law, where she graduated third in her class. She is admitted to
practice in New York and the Philippines.
Epiq AI Labs, established by Epiq in 2025, brings together builders, researchers,
academics, legal professionals, and industry leaders to advance AI for complex legal
and corporate challenges. Its work includes foundational AI research, large-scale data
analysis, collaborative development of practical solutions, and AI education. Alongside
research, the Labs leads the development of Epiq’s flagship AI platform – the foundation
for AI-powered workflows across litigation, compliance, antitrust, and other complex
legal and corporate matters.
Legal AI is often evaluated backwards: we see an impressive tool and only afterward
ask whether its work is reliable. But lawyers already possess the raw material for a
better test, their matters, decisions, and professional judgment. This talk shows how
legal teams can turn that expertise into practical benchmarks, examine what AI gets
wrong, and decide which work can be safely delegated. Before asking what AI can do,
define what good legal work looks like—and make the technology prove it.
