Mitratech Buys BotDojo to Put AI Agents to Work in Legal Software
The deal is meant to let agents handle assigned legal tasks using matter, spending and document records. Two-way connections to outside AI tools are expected this fall.
Listen to this story
The audio brief
Story brief
3 key pointsMitratech plans to fold BotDojo’s agent technology into ARIES AI, positioning agents to work directly with legal matter, spending and document records rather than relying on users to provide context in a chat. The approach pairs task execution with permission controls and human review, while planned two-way MCP connections would link Mitratech systems with outside AI tools. That connectivity is not expected until...
- 01
Planned agents target intake sorting, invoice review, contract turnaround and matter reporting, with supporting context attached to completed work.
- 02
Mitratech describes allow, deny, approval-required and expiring-access controls for users, agents and connected tools.
- 03
The company cites 50-plus connectors and relationships with more than 12,000 legal teams; these figures do not indicate BotDojo agent adoption.
Mitratech said on September 28 that it has acquired AI startup BotDojo to bring task-handling agents into the software legal teams use to manage their work. The goal is to give those agents access to the records behind an assignment, while keeping their actions subject to permissions and human review where judgment matters.
From stored records to assigned work
Mitratech is adding BotDojo’s agent technology to ARIES AI, its existing AI initiative. It plans to place agents inside its legal systems of record: software that holds information about matters, spending, documents and past work. That gives an agent a proposed route to the details needed to complete a task, rather than asking a user to supply them in a blank chat window.
The company describes agents prepared for specific jobs, including sorting incoming requests, reviewing invoices, turning around contracts and reporting on matters. Agents are meant to pick up assignments from queues, schedules and triggers, with a human handoff where judgment matters. Mitratech says they would return their work with supporting context attached, giving a reviewer more than a bare answer.
A connection in both directions
Mitratech also plans two-way support for Model Context Protocol, or MCP, a way for AI tools to connect to outside software. One direction would let external assistants use permitted context from Mitratech’s matter, spending and document systems. The other would let Mitratech agents reach tools a legal team already uses. The company expects that two-way support to be live in fall 2026; it is not presenting the full connection as available today.
That distinction matters for a team already using another AI assistant: Mitratech’s plan would not require every request to begin inside Mitratech. It would, however, put the company’s permissions between an outside tool and legal records. Mitratech describes allow, deny and require-approval controls, along with access grants that expire. Those are the safeguards it says will govern what users, agents and connected tools can see and do.
What the acquisition does not settle
Mitratech says it has more than 50 pre-built connectors and relationships with more than 12,000 legal teams worldwide. Those figures describe its potential reach, not how many teams use BotDojo-powered agents. The company says conference attendees in October will see live MCP connections, integrations and agent workflows. That demonstration should make the proposed handoffs more concrete, but it will not by itself show how agents perform across customers’ legal work.
The announcement calls BotDojo an acquisition but gives no purchase price or closing date. For legal teams, the more immediate open question is how much work an agent can complete before a person needs to step in. Mitratech’s design explicitly provides for that handoff; its announcement does not measure how often one will be needed.
Sources
- globenewswire.comMitratech Fast-Tracks Autonomous Legal Execution with New AI Acquisition
Reader comments
Newest comments first. Replies stay oldest first.