Deel's Akai agent learned by recording employee screens, automated 3,000+ hours of finance work
JafarNajafov · x · 2026-09-22
While hitting $140M ARR in 90 days, Deel quietly built Akai, an AI agent trained by recording employees' screens as they worked. It has already automated 3,000+ hours of month-end finance work — matching bank transactions against NetSuite, pulling data across systems, screenshotting for auditors, chasing missing info across teams, and submitting paperwork into FloQast across 100+ finance managers and 300+ entities. An employee shows it the process once; it captures the steps into a reusable workflow, stores new edge cases as they appear, and can share the workflow with a colleague abroad. Deel is now opening it to other companies.
Related event: Deel Launches Akai, Its Internal AI Agent Platform(2 posts)→
More from coding & agent
- jev-model-router Routes Claude Code Tasks Between Models via Function Hooks — dr_cintas · 2026-09-22
- A coding agent prompt: use my dotfile, keep keys safe, test locally without help — yoobinray · 2026-09-22
- Why devs skip Tinker for posttraining: cost and owning your own code — silver__tsuki · 2026-09-22
- codebase-memory-mcp indexes codebases into a persistent knowledge graph in milliseconds — tom_doerr · 2026-09-22
- Ambion: A Collaboration Kernel Running Multiple Agent SDKs in One Shared Room — andreisavu · 2026-09-22
- Formal Verification and Chaos Testing for Multi-Agent Systems with Dafny — andreisavu · 2026-09-22