One living asset: the Human Layer Graph.
A living, queryable model of how your organisation actually works: every role, person, skill, task, tool and process, and how they all connect. Not an org chart. Your workforce intelligence layer.
Every object in your organisation, connected.
Not a list of records, but the dependencies between them. Eight kinds of object, one connected model of how people, work and stack fit together.
Roles
Every role, decomposed into the tasks it actually does, not the JD it was hired against.
People
Who holds the work today, the skills they carry, and where their capacity is going.
Tasks
The atomic unit of the graph. Each one scored for automation potential, with evidence.
Skills
The skills behind every task, with the hours behind each, projected out 6 to 24 months.
Processes
How work actually flows across teams and systems, not how the org chart says it should.
Tools
Your stack, mapped to the tasks and teams that use it, and what AI has made redundant.
Vendors
Every vendor relationship, tied to the work it supports and the spend it carries.
Agents
The AI agents at work in your org, the tasks they have taken on, and what they freed.
Deeper every cycle. Sharper every decision.
Every programme, signal and result feeds back in. The longer it runs, the more the value compounds, and it stays yours.
Maps capability, not hierarchy
Every role carries rich context: the tasks it does, tools it uses, skills it needs, and how automatable the work is.
Learns from every interaction
Corrections, agent conversations and analysis runs feed back as memories. It gets smarter the more you use it.
Answers questions no other system can
“What happens if we sunset Salesforce?” “Where are our biggest automation opportunities?” Connected, analysis-backed answers.
Becomes your AI infrastructure
Expose the graph to every internal agent you build, so each one reasons over a structured model of how your workforce actually operates.
Open to every agent you trust.
The graph is not locked inside the platform. Query it from your own agents and tooling, with your governance boundaries intact.
Make the graph your AI infrastructure.
Expose the Human Layer Graph over MCP, a REST API, or the CLI, and every agent you build can query a rich, structured, analysis-backed model of how your organisation actually works.
Claude Code
Give your coding agent full context on every role, system and dependency as it works.
Claude cowork
Let cowork agents reason over your live org: plan headcount, model changes, draft JDs.
Any MCP client
Cursor, Cline, your own internal agents: anything that speaks MCP can query the graph.
Read more about the interface on Agents, MCP and CLI.
Build the foundation before you build on it.
Start with Discovery. We baseline your workforce task by task and stand up your graph.