Planhat news and product updates

Pi on Planhat OS can now search the public web and draft and send emails from within a conversation. Users can research an account before a call and have Pi send a follow up. Admins choose whether each tool is available through Pi's tool controls.

Pi moves from drafting email to sending it for a user, and admins decide tool by tool whether web search and sending are available.

Admins on Planhat OS can now restrict what data the Pi agent can touch, down to individual models and properties such as end user records or a sensitive custom field, so Pi cannot read, create or update them. Admins can also exclude specific companies and all their related data from Pi's processing.

Customer success teams can roll out an agent across the account base while keeping named customers and sensitive fields out of its reach, a common condition in enterprise contracts.

Planhat introduced Pi (Planhat Intelligence), an AI agent inside the new Planhat OS that reads customer emails, calls, tickets and usage data, answers questions in chat Sessions, builds rules based Agents and HTML dashboard pages, and raises Signals such as churn risk and expansion opportunities. Pi is rolling out to a limited set of users first, with availability for all Planhat customers stated for the following month, and it can also be reached through Planhat's MCP server and API.

Revenue and customer success teams on Planhat get an agent that can draft replies, update records and run renewal or onboarding processes, with Planhat stating that Pi routes decisions for human approval under existing permissions. Buyers should confirm rollout timing for their tenant and which actions Pi may take without approval.

Planhat shipped four Automation releases between 9 and 14 September. Workflow Mode creates a record for every Automation run, so teams can inspect sequence progress, email engagement and outcomes in dashboards. The others add permitted sending hours and days for email steps, an AI shared inbox template that triages inquiries and drafts replies, and an AI outreach template that matches underused features to customer objectives.

A record per run makes automated customer outreach auditable one account at a time, which is what a customer success leader needs before trusting it. Sending windows are the small control that keeps an automation from emailing a customer at three in the morning.