Want to connect with Mary Technology?
Join organizations building the agentic web. Get introductions, share updates, and shape the future of .agent.
Is this your company?
Claim this profile to update your info, add products, and connect with the community.
Mary Technology is active in the specialized data extraction and knowledge management layer of the AI agent stack. While many general-purpose agents struggle with the precision required for legal work, Mary functions as a high-fidelity extraction agent that converts unstructured case files into a structured state. This structured output—the "fact object"—is essential for building more complex agentic workflows in the legal industry, such as automated brief writing or strategy assessment, which require a verified ground truth to operate reliably.
For the broader ecosystem, Mary represents a move toward vertical AI where the agent's primary value is not in its chat interface but in its ability to enforce data integrity. By linking every extracted fact to a specific source and detecting metadata conflicts, they provide a blueprint for how agents can be deployed in high-consequence environments where hallucinations are not an option. They are effectively building the structured knowledge graph that future legal agents will use as their primary source of truth.
Mary Technology is an Australian legal tech company that builds what it calls a Fact Management System. Founded in 2023, the startup addresses the problem of "fact chaos"—the labor-intensive process of manually identifying, verifying, and organizing thousands of disparate facts from thousands of documents during litigation. While much of the legal AI market focuses on broad document review or chat-based research interfaces, Mary is built for the specific task of building a case chronology and fact record.
The technical architecture of the platform is what separates it from the current wave of legal AI tools. Most unified AI platforms rely on vector databases and Retrieval-Augmented Generation (RAG) to find relevant snippets of text. Mary, instead, extracts facts into discrete data objects. Each "fact object" is enriched with metadata regarding the people involved, the date of the event, and potential conflicts or inconsistencies with other evidence in the record. Because every fact is linked directly to its original source document, the system allows lawyers to audit the AI output without manual re-reading. This focus on structured data over conversational retrieval is designed to meet the "zero fault tolerance" requirements of high-stakes litigation.
Originally founded in Sydney by Daniel Lord-Doyle, a former lawyer, along with technologists Rowan McNamee, Harry Raworth, and Luke Abagi, the company initially served large enterprise clients and commercial litigation firms. Early adopters include Australian firms such as Shine Lawyers, Maurice Blackburn, and Hall & Wilcox.
In early 2026, the company closed a A$7 million (approximately $4.9 million USD) seed round led by OIF Ventures to fuel its expansion into the United States. This growth phase includes the opening of a San Francisco office and the launch of a self-serve platform. This new offering is intended to lower the barrier to entry for small to mid-sized law firms that lack the massive back-office resources of international enterprise firms but face the same document-heavy discovery challenges. By productizing the extraction workflow, Mary aims to reduce the time spent on fact management by 50% to 90%.
Mary Technology sits in a specialized niche of the legal agent stack. It is not a generalist assistant but a dedicated extraction agent. Its competition includes established legal tech giants and newer AI-native entrants like Harvey, yet Mary’s emphasis on the "fact record" rather than the "document search" provides a more granular utility for trial preparation. The company has also participated in major industry accelerators, including the AWS Global Generative AI Startup Accelerator and the A&O Shearman FUSE cohort, signaling strong institutional support from both cloud providers and law firms. As litigation continues to become more data-dense, Mary's focus on turning unstructured document folders into structured fact intelligence puts them at the center of the push for verifiable AI in the legal profession.
An AI-powered system that reviews thousands of documents to extract key facts, dates, and people into a verifiable record.
better Drawer primitive for Base UI
Arch repackage for Kagi OrionGTK
⚛️ Fast 3kB React alternative with the same modern API. Components & Virtual DOM.
a fast, modern browser for the npm registry
Find the cost of adding an npm package to your app's bundle size
work-in-progress atproto PDS
The Bluesky Social application for Web, iOS, and Android
Statusphere, but in atcute and SvelteKit
Mary Technology is hiring
You've explored Mary Technology.
Join organizations building the agentic web.