Mazle AI
building the intelligence layer for hiring. mazle auto-captures structured feedback from every interview, contradiction-checks it against the transcript, and compounds it into org-wide hiring intelligence.

founder
$ whoami
every interview generates signal.
most of it evaporates. mazle catches it.
building the intelligence layer for hiring. mazle auto-captures structured feedback from every interview, contradiction-checks it against the transcript, and compounds it into org-wide hiring intelligence.
led the ai product roadmap for samvaadini, a voice-bot platform for blue-collar recruitment in india.
built data models for customer service on medical imaging devices (mri, ct). worked with a distributed team across india, the us, the netherlands, and germany.
in 2020 i had an amazon internship offer and a full-time role at philips. i took philips. out of fear i'd fail the amazon bar. the role turned out to be nothing like what was promised, and then covid shut every door to switch. i spent a year stuck, applying for dev roles and failing at every one.
i stumbled into product management by accident. it clicked. three years later i was running ai product at vahan, and i saw up close how much signal gets lost between an interview ending and a hiring decision getting made. debriefs were chaos. feedback came days late. the best interviewer in the room often forgot what they'd heard.
mazle is my attempt to fix that. build the infrastructure that makes every hiring team as good as the best recruiting agency. turn interviews into data that compounds.
the worst case is i go back to being employed. that's not a bad worst case.