Revenue teams
Your CRM can't see product usage, support history, or what engineering shipped for a customer.
+Agent-native CRM
+Visitor ID & lead scoring
+Sequences & routing
- hubspot
- attio
- gtm point tools
SYS/01 — Open-source company OS
Replace the stack. Keep the context.
Aexy is an open-source company OS — CRM, sprints, workflows, docs, and people in one system your team, your AI agents, and the assistant you already use can actually share. Self-host it, or let us run it.
- hubspot - jira - notion - zapier
+ aexy
FIG. 01 — Operating graph
Synced
Engineering
24 active tasks
Sprints, releases, velocity
GTM
18 hot accounts
Visitor ID, scoring, routing
People
7 growth plans
Hiring, reviews, learning
Knowledge
1,284 indexed docs
Docs, graph, MCP tools
Sales agent routed a high-intent account
policy checked · CRM enriched · task created · Slack notified
SYS/02 — Platform
Every team's system. One place.
Aexy is not another point solution. It connects the systems companies normally buy separately, then gives AI agents governed access to that shared context.
FIG. 03 — CRM record, agent-enrichedLive
+Schema-less CRM
+Visitor ID & lead scoring
+Sequences & routing
+Customer health
SYS/03 — Who it's for
Start where it hurts most.
You don't adopt a company OS in one day. Start with the workflow your team is fighting today, then expand into the shared operating layer.
Engineering teams
Sprints, tickets, and releases live in trackers that know nothing about customers or revenue.
+Sprint lifecycle & tasks
+GitHub sync & analytics
+Release readiness
- jira
- linear
- stray trackers
Founders & operations
Docs, workflows, hiring, and reporting are scattered across a dozen subscriptions nobody reconciles.
+Docs & knowledge graph
+Workflows & approvals
+Hiring & reviews
- notion
- zapier
- hr point tools
SYS/04 — AI agents
Agents with real access — and a paper trail.
Aexy agents can read CRM history, draft emails, enrich accounts, update records, call workflows, and escalate to humans through policy gates.
LOG/01
Tool access
CRM, email, enrichment, Slack, workflows, docs, and company records.
LOG/02
Policy gates
Require approval, block tools, restrict fields, rate-limit actions, and cap spend.
LOG/03
Automation hooks
Invoke agents when a lead replies, a deal changes, a ticket arrives, or a workflow branches.
LOG/04
Audit history
Every run, tool call, policy decision, and config change is visible.
SYS/05 — Model Context Protocol
Already using ChatGPT? Point it at your company.
Aexy speaks MCP, so the assistant you already pay for can work inside your workspace — moving tickets, updating deals, reading sprints, drafting from real context — instead of you copying it into a chat window and pasting the answer back.
The assistant you already have
ChatGPT connects over HTTP. Claude Code, Claude Desktop, Cursor, VS Code and Codex run the server locally. Nothing new to learn.
It does the work, not just the reading
Move a task to In Review and assign it. Create a deal. Log a ticket. Query analytics. Not a chatbot bolted onto a search box.
Sign in, don't paste a key
ChatGPT registers itself and sends you to a consent screen where you pick one workspace. No API key to mint, store, or forget about.
You can take it back
Every connector is listed with the workspace it reaches and when it last ran. Revoking cuts it off immediately — it cannot quietly reconnect.
Your assistant
Connected to Aexy over MCP
Connected
Move the auth refresh bug to In Review and assign it to Priya.
aexy_sprints · bulk_update_statusaexy_sprints · assign_task
Done — moved to In Review and assigned to Priya.
Sprint 24 boardupdated just now
Todo
Rate-limit the webhook retry
In Progress
Backfill workspace slugs
In Review
Auth refresh drops the session
PPriya
SYS/06 — Open source
Nothing to hide. Read the source.
Aexy is open source and self-hostable, with exportable data and auditable logic. Use cloud to move fast or run it on your own infrastructure.
Self-host free
Run the full OS on your own infrastructure at no cost.
AGPL-3.0 licensed
The entire codebase is public — read how every feature works.
Data export
Your records, docs, and workflows are exportable, always.
Commercial cloud
Managed hosting when you want speed over ops.
FIG. 02 — Your infrastructureAGPL-3.0
$ git clone https://github.com/aexy-io/aexy
$ cd aexy && docker-compose up -d
# backend :8000 · frontend :3000 · temporal :8080
SYS/07 — Compare
Side-by-side with the tool you use today.
Aexy replaces point tools one workflow at a time. See exactly what you gain — and what changes — against the tool your team uses today.
Aexy vs Jira
Project tracking that also understands customers, docs, and AI agents.
Aexy vs Linear
Fast issue tracking, plus the rest of the company OS around it.
Aexy vs HubSpot
CRM and GTM without per-seat sprawl — open source and agent-native.
Aexy vs Salesforce
CRM depth for growing teams without the implementation tax.
Aexy vs Attio
A flexible CRM data model, connected to engineering and workflows.
Aexy vs Notion
Docs and knowledge with real structure — plus workflows that act.
SYS/08 — FAQ
Before you replace the stack.
What is an AI company operating system?
An AI company operating system is one workspace where core company data, workflows, and AI agents share context across teams instead of living in disconnected SaaS tools.
Can Aexy replace our CRM?
Aexy includes a custom-object CRM with contacts, companies, deals, activities, email sync, automations, and GTM intelligence. Teams can start with CRM and expand into engineering, docs, workflows, and people operations.
Can Aexy be self-hosted?
Yes. Aexy is open source and self-hostable, with a cloud option for teams that want managed infrastructure.
How do Aexy AI agents work?
Aexy agents run inside governed company context. They can use approved tools such as CRM records, email, enrichment, Slack, workflows, and docs, with policy gates, approvals, and audit history.
Can I use Aexy from ChatGPT or Claude?
Yes. Aexy ships an MCP server, so ChatGPT, Claude Desktop, Claude Code, Cursor, VS Code and Codex can work inside your workspace directly — reading sprints, moving tickets, updating CRM records, querying analytics. ChatGPT connects over a remote HTTP endpoint with OAuth: it registers itself, you approve one workspace on a consent screen, and the access it gets is scoped to you in that workspace and revocable at any time under Settings → Connected Apps.
How is Aexy different from Jira or Linear for engineering teams?
Jira and Linear track issues in isolation. Aexy covers sprints, tasks, GitHub sync, and delivery analytics — connected to CRM, docs, and workflows in the same workspace, so planning reflects customer commitments and AI agents can act across all of it.
How does Aexy compare to HubSpot or Attio for revenue teams?
Like HubSpot and Attio, Aexy includes a schema-flexible CRM with visitor identification, lead scoring, sequences, and routing. Unlike them, it is open source, self-hostable, and agent-native — and the CRM shares context with engineering and operations instead of living in a silo.
Adopt it one workflow at a time.
Bring engineering, GTM, people, knowledge, and AI agents into one system of record — starting with the workflow that hurts today.