Introducing Agent OS: Customer Intelligence That Starts Itself

Customer centricity always had a catch.

Most teams strive to put the customer first. But often, but they don’t have the infrastructure to consistently turn feedback into decisions, action, and accountability at scale.

Often, things get missed because they needed someone with the time, bandwidth and energy to drive it forward: running a report, investigating a question, creating a doc or ticket, communicating a proposal.

We’re launching Agent OS to change this, with new capabilities that teams can use to start and perform agentic work proactively.

Imagine this: help docs that rewrite themselves every week, that analyze the most common recurring themes and the questions your docs answer poorly, and fix them before the next wave of tickets. Customers get a better experience, support load improves, and nobody needs to remember to take a look.

That's the seat at the table customers never had, thanks to Agent OS — the operating system for customer intelligence.

Agent OS brings together two products in Enterpret:

• Enterpret Agent: the AI agent users can collaborate with in real time. A supercharged evolution of our Wisdom AI that now does much more, thanks to the ability to use skills, connectors, tools, and create artifacts.
• Agent Automations: create repeatable jobs that run the background — either on a schedule or in response to signals that you define

Meet Enterpret Agent: Wisdom, evolved

Enterpret Agent is the new way teams work with customer feedback in real time. It’s the supercharged evolution of Wisdom that expands what an AI assistant can do inside Enterpret.

At its core, Enterpret Agent sits on top of the Customer Context Graph. You converse with it, steer its investigation, and collaborate on outputs.

With Enterpret Agent, teams can:

• Ask rich questions about customer feedback and get evidence-backed answers.
• Create and use Skills: packaged workflows for repeatable work like VoC reporting, launch analysis, deep research, QBR prep, and escalation triage.
• Use Connectors to natively bring in third-party context at query time — product analytics from PostHog and Amplitude; CRM data from Salesforce; issues from GitHub, Linear, and Jira; logs from Sentry; docs from Notion and Canva; notes from Granola and more.
• Take action in connected tools — create a Linear ticket, draft a customer email, post to Slack, write back to Salesforce, etc.
• Create durable artifacts inside the conversation — charts, dashboards, reports, feeds, collections, and other objects the whole workspace can share and reuse.
• Benefit from agent memory over time, so recurring focus areas, preferences, and organizational context don’t need to be re-explained every session.

Introducing Agent Automations

Agent Automations are repeatable workflows, powered by Enterpret Agent, that you can build within Agent OS. You define what it should do, when it should run, what inputs it can use (data, tools and connectors), and where its output should go.

Uniquely, they can be triggered in two ways:

• On a schedule (daily, weekly, etc.)
• Based on a signal (runs when a signal is detected - like a feedback spike or trend)

Within Agent OS, you’ll find various templates included as examples you can build from.

Scheduled automation templates:

• Help Center Gaps — Improve help docs by tracking recurring questions not covered by existing content and auto-drafting improvements.
• Account Health Monitor — Monitor churn risk, expansion signals, sentiment shifts, and themes, and draft action plans for CS teams
• Launch Monitor — Track how a new feature lands after release each day: confusion, praise, bugs, adoption blockers, and themes, and auto-create tickets for issues
• VoC Report — Report on top complaints, requests, praises, and competitive signals each week, with week-over-week movers.
• Support Team Scorecard — Track agent-level CSAT, reply times, and quality trends, and draft coaching opportunities for managers.
• Customer Story Tracker — Identify and auto-draft new case studies and testimonial-ready quotes for marketing teams.
• Wall of Love — Find top customer praise, ready to share with the team in Slack.

Signal-driven automation templates:

• Quality Monitor — Detect feedback spikes, identifies likely root cause, shows who is affected, and creates the ticket or routes the issue for action.
• Escalation Shield — Identify urgent or emotionally charged customer feedback, especially from high-value accounts, and alerts the right teams early.
• Deal Risk — Detects when an active sales opportunity is trending toward loss and notifies the account owner and manager with evidence (coming soon).
• Sentiment Shift Detection — Surfaces meaningful shifts in customer sentiment across support, social, reviews, surveys, and other sources (coming soon).

Fully customizable for your unique workflows

Each agent automation can be easily customized with a prompt:

• "Every Monday at 9am, summarize the top complaints and improvement requests from enterprise customers, grouped by product area, and post in #enterprise-product."
• "After every launch, monitor feedback from the first 14 days, compare it to the previous release, and create a doc with adoption signals, blockers, and recommended fixes."
• "When emotionally charged feedback comes from a customer above a specified ARR threshold, notify the account team and draft a response with supporting evidence."
• "Every Friday at 11am, review support conversations for coaching opportunities and summarize patterns for the CX leadership team."

Available in beta today for all customers

Everything within Agent OS is available to all customers today, and free to use for the next 90 days!

Log in to get started, or check out the help center for more detail.