A PushMetrics agent at the center, surrounded by the things it can do: SQL, charts, email, Slack, Tableau and memory

An AI Agent is a chat-based assistant that lives inside PushMetrics. You ask it questions in plain English and it does the work for you: writing SQL, building charts, drafting an email, posting to Slack, exporting a Tableau view, or running a report. Anything you can do by clicking around in PushMetrics, an agent can do for you in a conversation.

Agents are different from a normal automation. You don't have to build a flow with steps and triggers. You just describe what you want. The agent picks the right tools and gets it done.

What can agents do?

Query your databases
Ask a question and the agent writes the SQL, runs it, and shows you the results.
Build charts
Turn results into interactive charts you can hover, zoom, and share.
Send emails
Draft and send emails with tables, charts, and attachments to the right people.
Post to Slack
Send updates, alerts, and reports straight into a channel or thread.
Export from Tableau
Find a workbook, apply filters, and export the view you need.
Remember and learn
Save notes, follow team playbooks, and get better with every conversation.

How it works

1. YOU ASK "Show me sales by region" 2. AGENT PLANS Picks tools and writes SQL 3. TOOLS RUN SQL runs, chart renders 4. RESULTS Table and chart appear in chat

The agent uses the databases, integrations, and contacts you've already set up in your workspace. Workspace admins decide which of those each agent is allowed to touch.

Everything happens live. You see the SQL being written, the chart being drawn, and the message being sent as it happens.

Where you can talk to an agent

You don't have to be in PushMetrics to use an agent. The same agent works from a few different places, and every conversation is saved so you can come back to it later.

In the app
Open the Requests page in PushMetrics. Start a new conversation or pick up an old one.
From Slack
@-mention the agent in any connected channel. The reply comes back in the thread.
On a schedule
Run an agent automatically. For example, send the weekly pipeline report every Monday at 9am.
Replay a past one
Re-run an old conversation in safe mode. The agent goes through the steps without sending real emails or making real changes.

Stays inside the lines

Agents are powerful, so PushMetrics gives you a few ways to keep them under control.

Spending limit per conversation
A single conversation stops if it goes over budget. No runaway costs.
Monthly budget per agent
Set a 30-day cap. The agent stops accepting new work once it hits the cap.
Per-tool permissions
You decide which tools each agent can use. Nothing happens by accident.
Full history
Every message, every tool call, every result is saved with the conversation.

Getting started

🚀
Every PushMetrics workspace comes with a default agent called PushMetrics Assistant. It's already set up with the most common tools and permissions, so you can just open the Requests page and start asking questions.

When you want to make your own agent, see Configuring Agents.

A quick glossary

A few words you'll see across these docs:

Agent. An AI assistant you've named and configured. It has its own personality, model, and set of tools.
Conversation (or ticket). One chat with an agent. PushMetrics calls these tickets because each one has a status, an assignee, and a full history, like a help-desk ticket.
Tool. One thing an agent can do, like run a SQL query, send an email, or post to Slack. You'll see each tool call as a card in the chat.
Skill. A short playbook you write once. The agent reads it and follows the steps. For example, "how to send the weekly pipeline report".
Memory and Knowledge Base. Memory is the agent's own notebook (one per agent). The Knowledge Base is a shared library every agent can read.