Platform
Opafex
The Agent, the Chatbot and the Operatives are three surfaces onto one system.
Five interconnected systems, one self-improving execution loop.
One system underneath
Three ways in, one memory
Each product is a surface onto the same platform. Teach it through one and the others already know.
Connected knowledge
Everything it learns becomes connected knowledge
Documents, conversations, outcomes and connections file into one grid — and the agent retrieves exactly what’s relevant, exactly when it needs it.
Every Opafex agent starts here.
One intelligence. Your business.
Five systems, fully interconnected.
Each one makes the others smarter.
A continuous execution loop.
Plan, approve, execute. Learn and repeat.
Your agent learns by doing.
Every interaction builds structural understanding of your business.
You upload documents and connect your CRM. Your agent knows what you gave it and nothing more.
The same system. The same five parts.
You choose how much control to keep.
Risk and safety class gate every action. Low-risk work runs; sensitive actions wait for you.
Your agent builds each of these.
It delivers results.
The Agent is live in beta from £500/month, usage included.
Get started0
OAuth providers
0
Operations
0
Universal connectors
0
Knowledge domains
Ask in plain English. The agent builds and acts across your tools and your team.
It designs your dashboards
Describe what to track and the agent builds the dashboard live. Widgets bind to real data or a table you fill in.
It builds you a database
The agent builds durable tables for a CRM or anything else. Fill them by CSV and turn them into widgets and reports.
It runs your field team
Operatives are no-code mobile apps that work with no signal. Photos, GPS, scans and forms sync back to your dashboards live.
It schedules and reacts
Say “every Monday at 9” and the agent schedules it without cron. Event triggers act when a row lands or a task finishes.
These are properties of the engine itself. It handles unsupported claims, dead processes, expired tokens and trivial steps.
Grounded by architecture
The engine separates tool calls from the answer and checks every citation. It re-reads every write; the agent cannot quietly invent facts.
Durable execution
Tasks run on a crash-safe engine with heartbeat recovery. If a process dies, the task resumes from its exact step.
Self-healing connections
The platform tracks connection health and ignores one-off hiccups. Only repeated failures drop a link; the agent carries on after recovery.
Cost-aware by step
Each step gets the right model for the job. Routine work uses a cheap model and hard steps a stronger one.
One agent is the unit of work here.
It is not the limit.
Every agent publishes an agent card that other agents can use as a tool. Delegation stays inside your tenant and the platform records every hand-off.
A second agent is another pair of hands. It shares your connections, your approval queue and the record of who allowed what.
All of this runs today. What comes next can wait until it ships.
The work it does
Three things most AI tools can’t — and that solve a real problem on day one.
Why you can trust it unattended
Four mechanisms, not four promises
We mean unattended literally. The code enforces each check and you can watch it run.
The platform blocks made-up answers.
The agent re-reads the database after every write. Execution gates catch fabricated IDs and force a grounded retry.
Built like a payments system, not a demo
The database is the source of truth, so nothing double-runs. A crash resumes from the exact step, proven 15/15 with zero stalls.
Nothing risky without a yes
One approval queue and an unbreakable safety floor no autonomy mode can cross. The trust profiler recommends more freedom — it never flips the switch for you.
It learns from outcomes
It improves by remembering what worked on the job. It never trains on your data, and you can read its notebook.
FAQ
Make it yours
You don’t need another dashboard.
You need an agent that works as hard as you do. Create your account and get started in minutes.