Describe the outcome. Your AI workforce organises itself.
Rebral1 plans the work, assembles the right agents, creates new specialists when a capability is missing, and returns finished work — with human oversight at every material step.
Runs on approved capabilities. Human approval on every material action.
rebral1.internal / p-101 / org-chart
EL
Engagement Lead
EL-01 · monitoring
MG
Manager / Orchestrator
MG-01 · ▸ running
RESEARCH TEAM
Market Research Analyst ▸
Document Analyst ✓
ANALYSIS TEAM
Data Analyst ▸
✦ NewFinancial Modelling Analyst
CONTENT TEAM
Report Writer ○
Presentation Designer ○
✦Financial Modelling Analyst created from approved capabilities.
How it works
Four steps, start to finished work
01
Submit an outcome
Describe what you need in plain language and attach any business context.
02
The Manager plans
Work is broken into tasks with skills, tools and dependencies identified.
03
The workforce assembles
Qualified agents are matched, and a new specialist is created from approved capabilities if one is missing.
04
Reviewed and returned
The Manager reviews and consolidates, humans approve what matters, and the deliverable is presented.
Capabilities
Built to be watched, not just trusted
Agent Org Chart
See your workforce as an organisation — Engagement Lead, Manager, and specialist teams — and watch it grow as new agents are created.
Dynamic Agent Factory
A missing capability becomes a new specialist, configured from approved templates only. Never from unrestricted tools.
Parallel and dependent execution
Independent work runs at once. Dependent work waits for exactly the output it needs, and receives it as a structured handoff.
Manager review and revision
Nothing reaches you unreviewed. The Manager approves, returns for revision, or escalates to a human.
Skill and Tool Registry
Agents can only ever use what you have approved. The registry is the hard boundary, enforced in data — not in a prompt.
Activity feed and audit log
Every decision, assignment, capability gap, agent creation and approval, recorded and exportable.
Governance
Autonomy with a hard boundary around it
Agents observe, recommend and draft. Anything material stops for a human.
Approved capabilities only
No self-granted permissions
Human approval on material actions
Retry, revision and creation limits
Validated structured outputs
Full activity and audit logging
FAQ
Questions worth asking
Research, analyse documents and data, draft content, create tasks, score and rank options, and produce reports — using only the skills and tools you have approved for that agent.
Sending email, any action with an external effect, enabling a capability that is not yet approved, and anything the Manager escalates after repeated failure or missing information.