Skip to main content
The AI Workbench is where Twin, your personal AI operator, works alongside you in the HR domain. Twin reads the same context graph you do, so its suggestions are grounded in real candidate history, offer precedents, and policy rules.

​
What Twin Does in HR

  • Context Assembly: when you open a candidate profile, Twin gathers related requisitions, past interviews, feedback themes, and comms history into a summary at the top of the page.
  • Offer Drafting: Twin can draft offer letters and compensation proposals based on the requisition level, location, and recent approved offers for similar roles.
  • Next-Step Proposals: Twin suggests what to do next, such as schedule a debrief, send a rejection, or escalate a stalled candidate, with reasoning tied to policy.
  • Delegation: Twin can hand off tasks to domain agents, such as asking the Scheduling Agent to find interview slots or the Compliance Agent to check policy attestation.

​
Example: Offer Draft

You open an approved requisition for a senior engineer in Austin. Twin surfaces the compensation band, recent offers for that level, and the approval chain. It drafts an offer letter with placeholders for candidate name and start date. You review, edit, and submit. Twin routes it for approval and tracks the DocuSign envelope.