Skip to content
evvolv.ai

About

We are building the team, not the tool.

Evvolv started from a simple observation: businesses do not lack software. They lack the hours to run it. So we built coworkers who do the running.

Software waits to be opened.

A coworker just gets on with it.

Every business we met had bought the software. Almost none had the hours to run it. So we stopped building tools and started building the people who use them.

The idea Evvolv was built on

Built by operators who ran systems at

  • Amazon
  • Google
  • Goldman Sachs
  • Carousell
  • J.P. Morgan
  • Unilever
  • Uber
  • Myntra
  • Truecaller
  • University of Pennsylvania
  • Mercedes-Benz
  • Nokia Bell Labs
  • Rakuten
  • Intel
  • OLX
  • Shuttl
  • Amazon
  • Google
  • Goldman Sachs
  • Carousell
  • J.P. Morgan
  • Unilever
  • Uber
  • Myntra
  • Truecaller
  • University of Pennsylvania
  • Mercedes-Benz
  • Nokia Bell Labs
  • Rakuten
  • Intel
  • OLX
  • Shuttl

The people behind it.

Operators first, founders second. Both spent a decade running the systems this replaces.

Rahul Gupta

Co-founder and CEO

Rahul Gupta

AI leadership across marketplaces, mobility and finance.

16+

Years in AI and data

IIT Delhi

Alumnus

  • Carousell
  • OLX Europe
  • Uber
  • Goldman Sachs
  • Myntra
  • Amazon
View full profile
Harshit Gupta

Co-founder

Harshit Gupta

The AI teams behind products used by hundreds of millions.

12+

Years in AI and data

GGSIPU

Alumnus

  • Truecaller
  • Rakuten
  • Unilever
  • Mercedes-Benz
  • Shuttl
View full profile

What we hold to.

Four positions that decide what we build and, more often, what we refuse to.

Do the work, do not describe it

Most software hands you a view of your problem and calls that progress. A coworker finishes the job and brings you the part only you can decide.

Earn the autonomy

Nothing leaves your business without a human yes. Autonomy is something an agent earns by being right repeatedly, not a switch we flip on day one.

Say when it fails

An agent that quietly guesses is worse than no agent. Ours stop, say what went wrong and hand back the context.

One coworker beats ten features

We would rather ship one specialist who genuinely owns a function than a menu of capabilities you have to assemble yourself.

Where we are going.

Two coworkers are working today: Vikram on growth and Maya on marketing. Six more are built and arriving through the year, each owning a function outright rather than assisting with it.

The end state is not a bigger product. It is a back office that runs itself, where the only thing left on your desk is the handful of decisions that genuinely need you.

Come and see it work.

Thirty minutes, your real data, no deck.