The recruiting agent built for job seekers in an AI-native economy.
Kevin Astuhuaman · Jasmine Rattan · Devansh Shah · UC Berkeley Haas MBA
People don't want to click through ten job boards anymore. They want an agent that watches the market for them and tells them what to do next.
Every tool in hiring was built for a world of Boolean search boxes. That world is ending.
of senior roles at top tech companies never appear on LinkedIn in the first 24 hours.
of listings on major boards are ghost jobs — already filled or never real.
The incumbents are paid by employers, not seekers. Their interface — keyword + location + title — can't express what you actually want.
Claude, ChatGPT, Cursor — people now expect tools to work conversationally and in the background. Job search is one of the last big markets still stuck in the old paradigm.
Whoever drives market liquidity in this new paradigm — matching seekers to jobs faster and better — wins the category.
Native apps for people who want a feed. MCP and CLI for people who already live inside Claude and ChatGPT.
weekly active users at Haas — half of our class, fully organic, no launch.
median detection lag vs. ~7.6 hours on LinkedIn in head-to-head tests.
Berkeley's Career Management Center started using our data this week. Students ask us to give it to friends and partners at other schools.
Start with job seekers at MBA and graduate programs. Expand into undergrad career centers. Then open the platform to every agent and company that needs real-time hiring data.
LinkedIn and Indeed are $25B+ businesses built for a search-box world. We're building the one for the agent world.