Abhas SinghFree GTM strategyLet's talk
Back to work
Case study / MahakramFounder & CEO · 2026–present

Research into the mind, built into a working system.

Mahakram grew from a question I kept returning to: why do people read the same situation differently, and how does that shape what they do next?

Live product

A classification engine planned and built solo

01 / The constraint

Self-description is not enough.

The research looks for observable signals in language, voice, and body language, then connects those signals to perception, judgment, and action.

02 / What I built

The work

  1. 01

    Built the research base

    Created a hand-built classification dataset and a multimodal analysis pipeline over text, voice, and video.

  2. 02

    Built the engine

    Used 768-dimensional embeddings and three rounds of elimination across 128 candidates, with confidence tracked by dimension.

  3. 03

    Made the output useful

    The system produces a cognitive profile and playbooks for pitching, hiring, negotiation, and management.

03 / Evidence

The work record

  • Live productUse the system and inspect its public research.
  • ResearchMethods and framing behind the classification work.
  • External outputA classification delivered to a founder from his own data and a video.