AIHR
AIHR is an AI-HR venture I’m building, still at concept stage. This page will carry the real story — what it does, who it’s for, and where it stands — once there’s something concrete to show. In the meantime, here’s what shapes the thinking behind it.
The vision
Engineering-first hiring, built beyond traditional ATS systems and keyword-matched, AI-generated CVs. Recruitment doesn’t need AI that just parses text faster — it needs AI that reasons about candidates the way a person does. I work with AI for HR applications every day at Knovik, and AIHR is where I apply that responsibly.
How I use EdgeTal as a pocket recruiter app
Most of my candidate search happens on my phone, between calls, not at a desk with a hiring dashboard open. EdgeTal runs entirely on-device, so I can pull it out mid-conversation and search the candidate pool in under a second, without waiting on a cloud round-trip or worrying about where that data goes.
That changes how I work day to day: a client mentions a requirement, I check the pool right there. A candidate’s profile shifts, I re-run the search before the call ends. It’s less “log into a system later” and more “ask the question as it comes up” — which is closer to how recruiting actually happens.
What Knovik engineering taught me about MCPs and HR automation
Working alongside Madusanka and Knovik’s engineering team on EdgeTal, I’ve learned to think about HR automation the way they think about MCPs (Model Context Protocol): as discrete tools an AI agent can call, not one system trying to do everything. Screening, scheduling, compliance checks — each is its own composable piece, not bundled into a single black box.
That framing is the seed of AIHR: HR automation built as a set of well-scoped tools a recruiter can trust individually, rather than one opaque AI making every decision at once.
My Research Experience
EdgeTal
Co-authored research on EdgeTal for EICON 2026. Built by the Knovik team; I contribute the practitioner-side research and evaluation.