The AgentivUse cases

Answer data access requests on time

An agent that logs each personal data request (access, correction, deletion), works out the deadline under the law that applies, searches Outlook, SharePoint, OneDrive and Teams for the person's data, and drafts the acknowledgement and the response for a person to review before anything is sent

Function: OperationsEffort: About a weekPublished: 2026-10-24
Claude CoworkMicrosoft 365OutlookSharePointMicrosoft Teams

The problem

When a customer or former employee writes 'send me all the data you hold on me' or 'delete my data', most small businesses have no process: nobody owns it, the deadline is unclear, and the data is spread across mailboxes, shared folders and chats. Missing the deadline or sending the wrong person's data is a bigger risk than the request itself.

How the agent works

  1. It starts

    The owner installs Anthropic's legal plugin in Claude Cowork or Claude Code and connects the Microsoft 365 connector, which needs a work or school account and can only search and read Outlook mail and calendar, SharePoint and OneDrive files, and Teams chats.

  2. The agent works

    When a request arrives, the agent identifies the type (access, correction, deletion, restriction, portability, objection or opt out), picks the regulation from where the person is and where the business operates, logs the date, type, requester, regulation, deadline and handler, and drafts an identity check proportionate to the risk. Once identity is confirmed it searches the Microsoft 365 sources for the person's name and email, lists what it found with links, and checks exemptions such as legal retention duties, litigation holds and third party rights. It drafts the acknowledgement and the response, or a partial denial that gives the legal basis and the right to complain to the regulator, from templates the business saves.

  3. The result

    Escalation triggers (minors, health or biometric data, a regulator or lawyer asking, an employee in a dispute) stop the template and send the matter to the owner or counsel, every draft is reviewed by a person before it is sent, and any deletion is done by hand in the source system.

Worked example

The compliance-check skill in Anthropic's legal plugin lists the deadlines it works with: GDPR and UK GDPR 30 days, extendable by 60 days with notice; CCPA and CPRA acknowledgment within 10 business days and a substantive response within 45 calendar days, extendable by 45; LGPD 15 days. Its legal-response skill holds request templates for acknowledgments, identity verification, fulfilment, partial or full denials and extension notices, tracks deadlines with calendar reminders, and says plainly that it does not provide legal advice. Neither skill covers Thailand's PDPA (the only PDPA it names is Singapore's), so a Thai business has to add its own rules to the templates. The Microsoft 365 connector page on claude.com lists only search and read tools (outlook_email_search, outlook_calendar_search, sharepoint_search, chat_message_search and others), so the agent can find a person's data but cannot change or delete it.

Data protection. This use case touches personal data, so check the rules where you operate (for example GDPR, PDPA). The agent reads exactly the data the request is about: one person's emails, chats and files, possibly including sensitive data. Thailand's PDPA gives people the right to access, correct and delete their data; check the PDPC's current rules on response times and request channels and put them into the templates before going live, keep the request log in a restricted location, send found data only to the verified requester, and check your AI vendor's processing terms.

Join the Agent OS Club

8 weeks to the AI team a 50 person company would hire. Instant access to our agent blueprints. USD 15 a month, cancel any time.

Join the Agent OS Club

Building this for an organization? Learn more at theagentiv.com

Sources

Related use cases

Every monday.com board clean by Friday

An operations agent that audits the monday.com boards every Friday for items with no owner, no due date, a stale status or a likely duplicate, posts the findings with a proposed fix for each, and applies only the fixes the project lead ticks, in bulk, through the official monday.com MCP

OperationsEffort: An afternoonmonday.com · Claude Code · Slack

A monday.com account that started with one clean board now has forty: items with no owner, due dates in the past, statuses nobody has touched since the kickoff, three columns that mean the same thing. The weekly report built on those boards is wrong before anyone reads it, and nobody has an afternoon free to tidy up.

Shopee stock that matches your warehouse

An agent that compares Shopee stock per variation with the warehouse sheet every morning, alerts the team on LINE about low or mismatched items, and stages stock updates for the owner to approve

OperationsEffort: About a weekShopee · Google Sheets · LINEThailand

A shop that sells on Shopee and from its own stockroom keeps two numbers for every variation: what Shopee thinks is left and what is on the shelf. When they drift apart the shop either oversells and has to cancel orders it cannot fill, or shows sold out on a product that is sitting in the back, and nobody notices until a customer complains.

Overdue tasks chased before the client asks

An agent that scans monday.com boards each morning for items past their due date, posts a nudge on each item and sends the project lead one summary

OperationsEffort: An afternoonmonday.com · Claude Code · Slack

Due dates on a monday.com board turn red quietly, and the project lead finds out when the client asks where the deliverable is. Chasing by hand means opening every board, scrolling for red dates and writing the same polite reminder ten times.