Most interview panels don't need more training. They need a better scorecard.
That's not what a client in Whitefield wanted to hear in 2021. They had run a two-day workshop for forty engineering managers, and six weeks later the same two managers were still 1.8 points apart on the same candidate. We asked to see the scorecard. It said "Culture Fit — 1 to 5." Nobody had ever defined what a 3 looked like.
Lensforge builds the thing that actually gets used at 11am on a Tuesday: a work sample that resembles the real job, an interview guide with behaviourally anchored ratings, and a psychometric section only where it earns its place. For an SDET role we'll write a flaky-test triage exercise drawn from your own repository's patterns. For a GCC finance team, a reconciliation case with a deliberately wrong number buried in it. Then we calibrate — two graders, one recording — and we don't hand it over until inter-rater agreement holds.
Most of our work sits along the Outer Ring Road corridor and the Manyata campuses, with some clients in Hyderabad and Pune. Content drops into whatever you already run — Workday or Greenhouse or SmartRecruiters. Candidate data handling is written against the DPDP Act rather than bolted on later, and accommodation variants under the RPwD Act are part of the build, not a form someone has to request.
Eleven of us. We take on fewer accounts than we could.
We're expanding delivery to US-based clients this year. See careers for the remote-contract role we're currently filling.
