Sales
Private AI workforce · invite-only preview
Meet Lucy, your AI workforce.
Message Lucy in Teams, Slack or from your phone. One message puts a team of AI specialists to work across your whole business.
Microsoft Teams
live
@Lucy what meetings do I have today?
Lucy is on it
LucyThree today: 9:30 site visit, 12:00 supplier call, 16:00 team catch-up. The 12:00 clashes with your dentist. Want me to move it?
Integrations
Plugged into what you already run.
Lucy works in your tools, not another dashboard.
-
Marketing
-
Sales
-
Finance
-
Support
-
Operations
-
Reporting
Each connection installs like an app, with no rebuild. Four are live today, the rest are built on request. See the full list.
The platform
Not a chatbot. A team for every part of the business.
Specialists who remember your business, and share what they know with each other.
Marketing
Campaigns written and scheduled.
Finance
Receipts read, variance explained.
Customers
Every enquiry answered.
Operations
Admin that runs itself.
Memory
They remember your business.
The idea
Hire the staff, skip the salaries.
Browse the catalog, hire who you need, give each a job description. Come back next week and they still know your voice, your customers, and what you decided last Tuesday.
It's like hiring staff without the salaries, the HR, or the meetings.
Collaborate
Ask the whole team. Watch them sort it out.
They hand work to the right specialist and come back with one answer.
Launch team · 4 agents
You
You
Here's the raw feature list for the new release. Plan the launch.
PM
Product Manager
Breaking this into positioning questions. Research, can you check how competitors describe similar features?
handed to Research: competitor positioning
Research
Three competitors lead on speed, one on privacy. Privacy is under-served, that's our gap.
M
Marketing
Angle: "the workforce that stays on your server." Content, draft the announcement around that.
C
Collator
Here is one launch document: positioning, the privacy angle, and a ready announcement post. Raw contributions are below if you want them.
Private
What one agent knows. Never visible to another. Enforced in code.
Team
Shared by one team. Projects in flight, decisions made together.
Company
Brand voice, customers, the facts. Set it once, and everyone you take on already knows it.
AI models
Pick a provider. Pick a level.
Or let Lucy choose for each one. You pay the AI provider directly for what you use, with no markup from us.
Provider
Level
Claude Sonnet
The daily default. Sharp, quick, and good with nuance.
- Reasoning
- Speed
- Cost
- ££
Questions
The honest answers.
Is this just one chatbot with extra steps?
No. The unit is a team, not a chat. Each agent has its own expertise, its own memory with real privacy boundaries, and its own tools, and they hand work to each other.
Where does my data live?
On your own server, in your own database, on your own filesystem. Single-tenant by design, so there is no shared database and no other customer on the machine.
Do you train on my data?
No. Your conversations and memory stay in your deployment. Models are called through your own API keys under their own terms.
Can one agent read another agent's private notes?
No. Private memory stays private. That boundary is enforced in code and covered by an automated test that fails the build if it ever leaks.
Can I add my own tools and integrations?
Yes. If a tool has a way to connect, we can connect it. New connections install from the catalog like an app, so nothing has to be rebuilt.
Request access
Hire your first team.
Invite-only while we set up the first deployments. There is no markup on AI usage: you pay the AI provider directly for what Lucy uses, and we set that up with you.