{"slug": "exclusive-minutebox-launches-ai-agent-to-turn-its-entity-management-database-a", "title": "MinuteBox Launches AI Agent to Turn Its Entity Management Database Into a ‘System of Work’", "summary": "MinuteBox launched the MinuteBox AI Agent, an optional per-seat feature that lets legal professionals use plain-language requests to retrieve corporate entity data, order reports from government business registries, draft documents from a firm's own precedents, and record corporate changes back to the entity's record. Cofounder and CEO Daniel Levine said the launch shifts MinuteBox from a \"system of record\" to a \"system of work,\" and predicted that \"within six months, 95% of our users will use the agent 95% of the time.\" The agent combines MinuteBox's structured entity database, a compliance engine that validates data against statutory and company-specific rules, and direct registry integrations into a closed loop, and is the first of three AI releases the company plans over the coming months.", "body_md": "In what its CEO says is a complete reshaping of the company and its product, entity-management company [MinuteBox](https://www.minutebox.com/) today launched a new AI agent that turns its corporate records database into a platform for performing legal work.\n\nThe new MinuteBox AI Agent lets legal professionals use plain-language requests to retrieve information about corporate entities, order reports from government business registries, draft documents using a firm’s own precedents, and record corporate changes back to the entity’s record.\n\nThe launch, which the company previewed exclusively for LawSites, is the first of three AI releases the company has planned over the coming months. Cofounder and CEO [Daniel Levine](https://www.linkedin.com/in/minuteboxdan/) said it marks the company’s shift from a database that stores corporate information to a platform where the work of corporate legal teams gets done.\n\n“MinuteBox has typically been a system of record,” Levine said in our interview. “What we’ve been working on over the course of the last year is turning it into a system of work.”\n\nThe agent draws on three features that were already in place within MinuteBox: a structured database of corporate entity information, a compliance engine that validates the data against statutory and company-specific rules, and direct integrations with corporate registries for searches and filings.\n\nBy bringing those together through an AI agent, Levine said, the result is a “closed loop” in which a user can pull a company’s information from a government registry, compare it to the firm’s own records, record a corporate change, generate the documents the change requires, and then file them with the registry, all without leaving the platform.\n\nAs of today, the agent is available to MinuteBox customers as an optional feature, priced per seat. The product is designed and priced to be used by paralegals and law clerks, as well as lawyers, Levine said.\n\nHeere is a video showing MinuteBox AI used to set up a Delaware subsidiary from its certificate. Other videos are [available on YouTube](https://www.youtube.com/@minutebox).\n\n### Chat Becomes the Starting Point\n\nOn Friday, Levine and two other MinuteBox founders – [Steven Pulver](https://www.linkedin.com/in/stevenpulver/), chief operating officer; and [Sean Bernstein](https://www.linkedin.com/in/sean-bernstein-bb509249/), president, Registry Services – gave me a demonstration. The fourth founder, chief technology officer [Brian Hunt](https://www.linkedin.com/in/brianmhunt/), was not present.\n\nUntil now, they explained, the starting point for MinuteBox users has been a dashboard listing every corporate entity they have permission to see. With the new agent, that starting point becomes a chat window – what Levine described as a single pane of glass into the corporate record database ecosystem.\n\n“I think within six months, 95% of our users will use the agent 95% of the time,” he said.\n\nPulver demonstrated this by typing a request for information on a sample Irish company, using an @ sign to select the entity from the database. The agent returned the company’s details, directors, officers and share capital.\n\nThe response was formatted in panels and tables, with graphical badges used to highlight jurisdictions. Users can expand and export any table, Pulver said, and the same kind of query can run across every entity in an account, such as a search for all companies where a named person is a director.\n\nThe agent can also perform tasks that have, until now, been done by other legal AI products. As an example, Pulver showed a tabular review of six commercial contracts across 11 columns, with each extracted value linked to its place in the source document and tagged with a confidence percentage.\n\nIn another example, the agent turned a share purchase agreement into a “visual dossier” – a summary page that a user could share with a client, showing the purchase price, closing date, escrow holdback and other key terms. Firms can customize this with their own logos, fonts and colors.\n\nEach agent conversation can also become a data room, or “AI room” as Pulver called it. With one click, a user can share the conversation and its files with a client or colleague, with controls for watermarks, expiration dates, PIN access, downloading and notifications. Clients can upload documents to the room, and multiple people can run their own chat sessions inside it.\n\nUsers can bring documents into a conversation by drag-and-drop or email, or from other integrated platforms. MinuteBox currently integrates with iManage, Clio Manage, OneDrive, Google Drive, Dropbox and Outlook.\n\n### Registry Searches From a Prompt\n\nMinuteBox already operates a registry services business that lets customers search and file with a broad array of government business registries. But while the typical process of conducting a corporate due diligence search can be cumbersome, the agent brings that into a natural-language workflow.\n\nIn our demonstration, Pulver asked the agent to order a corporate profile report for Pizza Pizza Limited, the Ontario restaurant chain. The agent searched the Ontario registry by name and returned a list of matches that included business-name registrations and older, inactive corporations. It identified the one active corporation with that exact name and then, because obtaining the report required payment, it paused for the user’s approval before placing the order.\n\nWhile the agent made it look easy, Levine said that most registries cannot easily be searched by a business name alone.\n\n“The agent takes that name, it sprawls out across different corporate registries and different name search registries to find a way to locate where that entity is located and come back with all the things that it found,” Levine said. “It then allows you to say, ‘Yes, I’m looking for the one in Ontario. Let’s dive deeper into that one.’”\n\nBernstein contrasted this with the typical manual workflow in a law firm, where the same task might involve three or four people, passing from a partner to a junior lawyer to a law clerk, and sometimes on to a clerk who specializes in searches. The task the agent finished almost instantly might have taken hours for a law firm.\n\nPulver next asked the agent to create an entity record from the profile report and to draft director consents using a Word precedent he attached. In what he characterized as a “happy accident,” the agent cut him off to inform him that his account already contains an entity that appeared to be the same business.\n\n“It’s telling me about a critical ambiguity, right,” Pulver said. “And this is super helpful for our end users, because in any AI use case, you want to be able to validate.”\n\n### Drafting From a Firm’s Own Precedents\n\nAt this point, Pulver had the agent draft the consents using the existing record, producing a consent to act for each director as a Word document, along with a table of the facts it had used.\n\nAlthough MinuteBox already has a document automation tool that combines templates with a no-code, drag-and-drop editor, the agent makes that even easier, matching any existing Word precedent to the entity data and generating a document without prompting.\n\nThis can be done for any documents, simple or complex, and firms can optionally add plain-language instructions to their precedents to control how the agent handles them.\n\nThe agent can also change the record itself. To show this, Pulver opened an individual entity’s record and instructed the agent to add someone as a director and officer and issue her 100 common shares.\n\nThe agent flagged two issues. It found that the account had more than one person with that name and asked which he meant. It also pointed out that the company had no share class called “common.” It offered to issue from the existing class or authorize a new one.\n\nAlthough not shown in our demonstration, Levine said the agent can carry a corporate change through to filing and write the results back to the record.\n\n### Built for Paralegals and Lawyers\n\nLevine said that MinuteBox has traditionally had two categories of users. One is the lawyers who use the product primarily to consume information. Maybe they need to check something quickly or look at the minute book.\n\nThe other category consists of the paralegals, clerks and other support staff who really go deep into the platform and use it to create work product.\n\n“No one is really building for the paralegal and for the clerk, certainly not on the corporate legal team,” Levine said. “We felt very strongly that we wanted to make sure that our product was built first and foremost for them.”\n\nHe said that MinuteBox has set its pricing so that firms could justify giving the agent to every paralegal – and even to mailroom staff who need to find the lawyer responsible for a corporate client.\n\n“It’s really important for us to make this accessible to the firm,” he said. “We want it to be a no-brainer for them to say, ‘Yeah, every paralegal should have this.’”\n\nAlthough he did not disclose the price, Levine said pricing is per seat and varies with several factors, including the volume of AI tokens a customer buys. “We expect that the monthly cost for a user will be around the cost of one billable hour for a paralegal.”\n\nCustomers can also bring their own key, connecting the agent to an AI model they already license instead of buying tokens from MinuteBox. Levine said that option offers substantially the same features, with differences that depend on the capabilities of the model they choose.\n\nWith today’s announcement, existing customers can decide whether to turn the agent on, and users can then choose between the dashboard and the agent as their starting point.\n\n### Unleashing a Workhorse\n\nFounded in 2017, MinuteBox grew out of its founders’ experiences practicing corporate law, where they repeatedly encountered the same problem. Every new transaction required another round of gathering and verifying basic information about the entities involved.\n\n“Every time we’re doing a transaction, it’s like starting from scratch,” Levine recalled.\n\nAlthough some law firms maintained corporate databases, those systems were often outdated and lacked modern automation capabilities, he said.\n\nMinuteBox set out to develop a modern, cloud-based alternative. In 2022, it raised a $4 million seed round to support its growth. Today, its customers range from organizations managing dozens of entities to one with more than 60,000.\n\nA key differentiator for the company, Levine said, has been its compliance engine, which checks corporate records against statutory requirements in different jurisdictions, as well as each company’s own governing documents.\n\nThat validation layer now provides part of the foundation for the AI agent, giving it access to structured entity information and rules against which to check proposed actions.\n\nMinuteBox’s agent launch is the first of three planned product announcements.\n\nThe company will soon release a model context protocol (MCP) integration that will let customers connect their own AI agents to the corporate information stored in MinuteBox.\n\nAs for the other announcement, the company is not yet releasing details.\n\nFor the immediate future, Levine said MinuteBox will focus on introducing customers to the broader capabilities of the agent and showing how it can transform their corporate legal workflows.\n\n“We’ve taken this system of record, which is supremely valuable,” Levine said, “and now we’re allowing it to be a workhorse.”", "url": "https://wpnews.pro/news/exclusive-minutebox-launches-ai-agent-to-turn-its-entity-management-database-a", "canonical_source": "https://www.lawnext.com/2026/11/exclusive-minutebox-launches-ai-agent-to-turn-its-entity-management-database-into-a-system-of-work.html", "published_at": "2026-11-16 16:59:10+00:00", "updated_at": "2026-11-16 17:23:41.893978+00:00", "lang": "en", "topics": ["ai-agents", "ai-products", "artificial-intelligence"], "entities": ["MinuteBox", "Daniel Levine", "Steven Pulver", "Sean Bernstein", "Brian Hunt", "LawSites"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/exclusive-minutebox-launches-ai-agent-to-turn-its-entity-management-database-a", "markdown": "https://wpnews.pro/news/exclusive-minutebox-launches-ai-agent-to-turn-its-entity-management-database-a.md", "text": "https://wpnews.pro/news/exclusive-minutebox-launches-ai-agent-to-turn-its-entity-management-database-a.txt", "jsonld": "https://wpnews.pro/news/exclusive-minutebox-launches-ai-agent-to-turn-its-entity-management-database-a.jsonld"}}