Everday, inside the AI you already use.
Claude, ChatGPT, Cursor, whatever your team runs: point it at your Everday workspace and it can search your talent graph, screen against your rubrics, draft follow-ups, read the evidence, with your permissions, under your sign-offs. That’s MCP: no new app to learn, your AI just knows your people-world.
claude · chatgpt · cursor · any mcp client
Ask about your people-world…
sample session, sample data · with your permissions, under your sign-offs
Concretely.
“Find internal candidates for this vacancy, and show the why per name.”
“Screen these ten CVs against rubric X.”
“Who in our talent base is one skill away from this role?”
“Draft follow-ups for everyone who stalled this week.”
drafts; you send“Show me the evidence behind yesterday’s shortlist.”
Your vocabulary travels.
Your org’s models and building blocks ride along as MCP resources: the AI speaks your workspace’s language, not generic HR. A “traject” stays a traject; your rubric stays your rubric.
riding along as resources · sample vocabulary
Same door, same rules.
Same permissions, same human sign-offs, same audit trail as every other door.
An agent with your key can do exactly what you can, and nothing more.
Platforms that trap your people-data in their UI are betting you’ll never want your own AI on your own data. We’re betting the opposite.
