Ulpianus
AI-powered legal research with verifiable sources.
The founders
The company
Ulpianus is a legal research platform for lawyers and law firms in Chile, built on an aggregated corpus of more than 6.5 million Chilean legal documents — Supreme Court and Courts of Appeals jurisprudence, current laws, doctrine and legislative history. Users query it in natural language to research case law, consult statutes, draft documents and analyse texts, with every answer carrying a citation that links back to the official source.
Generalist AI models are not trained on specific national jurisprudence, which makes them unreliable for Chilean practice and leaves a gap for a jurisdiction-specific tool. Lawyers now have the habit of working through chat interfaces, so distribution via the web, WhatsApp and assistants like ChatGPT, Claude and Copilot meets them where they already are.
Founded in 2025 in Santiago by Nicolás Barriga, the company says its aim is to make legal knowledge more accessible by applying AI to Chilean legal material, starting at home and with plans to expand across Latin America.
Details on the founding team's background are not public yet. What the company has put on the record is a set of product decisions that suggest a close reading of what law firms actually require: verifiable citations on every answer, client data excluded from model training, deletion after 30 days, and hosting in Latin America to comply with local data protection rules.
Who else is building in this space
Companies working on the same thing, at every stage.
Unicorn
AI platform for law firms and in-house legal teams
Unicorn
Agentic AI workspace automating legal review and drafting. | Formerly known as Leya
Unicorn
EDiscovery and litigation platform
Unicorn
AI-driven platform streamlining plaintiff law firm operations
Scale-up
Providing a generative AI solution for tax research
An AI Copilot for transactional lawyers for contract review
Thousands more founded every month.
Dealroom tracks new companies as they appear — founders, funding and signals, over the API or in the platform.
