vibencode
About

A team to work with.

We spent years building software for other people and got good at the unglamorous half: the integration, the edge cases, the handover. That is the half AI projects keep failing at, so it is the half we sell. The rest of our time goes to our own products and the people coming up behind us.

Building since 2021 Indore, Madhya Pradesh
What we are And what we are not

A deep-tech consulting and engineering practice. We advise, investigate, design and build, and we stay until the thing runs on a Monday without us.

Not a platform, not a licence, not a seat. There is nothing to subscribe to. You hire people, for a period, to do a defined piece of work, and you keep what they build. Most of them are startups, which is why nothing we leave behind assumes a platform team on the other side of it.

The founders Who you will actually meet
Co-founder

Prateek Porwal

Operations, talent, sales and business development. Usually the first person you talk to, and the one who scopes the engagement.

Co-founder & CEO

Samyak Jain

Product and delivery. Has spent his career turning ideas into things that actually ship, which is a narrower skill than it sounds.

Co-founder & CTO

Samarth Paliwal

Technology and engineering direction. Has been writing code since adolescence and still writes it every day.

How we got here The short version
  1. June 2021

    First international order

    Before there was a company there was a client, which is the right order for these things.

  2. May 2022

    First commercial delivery

    Shipped, in production, being used by people who were not us, which is the only definition of delivered we accept.

  3. February 2023

    Enough people, and then no more

    Enough of a team to run more than one engagement properly at a time, which is a different thing from running more than one at a time. We stopped there deliberately.

  4. 2024–2025

    The practice turns

    Full-stack and SaaS delivery was what we were good at. We kept the discipline and changed what sat inside it: retrieval, evaluation, agents in the loop, and the operational work that keeps any of it running.

  5. Now

    An AI consulting and engineering practice

    Advisory, applied research and hands-on engineering, and the same insistence that what we build has to survive being handed over.

How we work Six things we actually hold to
  • Say the uncomfortable thing earlyIf the use case is wrong, or the data is not there, you should hear it in week one rather than week nine.
  • Measure before buildingThe evaluation gets written before the interface. A capability nobody can assess is a liability.
  • Recommend against ourselvesSome problems do not need AI. Saying so costs us one project and earns the next three.
  • Design for the handoverWe work so that leaving is uneventful. Documentation, tests and a team that has been doing it alongside us.
  • Keep the human where it mattersApproval sits where money, data or reputation changes hands. Autonomy is earned, not assumed.
  • Ship, then widenBehind a flag, on real traffic, one team first. Big-bang launches are how you find out what you got wrong all at once.
Lean on purpose Why the team is the size it is

A small team is not a stage we are trying to grow out of.

Most of what goes wrong on one of these projects goes wrong in the gap between the person who understood the problem and the person who writes the code. Every layer you add widens it. The widest version is an agency with a sales team at one end and a delivery pod at the other, joined by a document.

So the person who scopes your engagement is on it. The founders are on it. Nobody is briefed through a specification, because there are few enough people involved that a specification is not the only way the understanding can travel.

It also means we can only sell what somebody here actually wants to work on. There is no bench to keep busy. In a field that moves at this speed that is not a constraint, it is the mechanism: the depth is a by-product of people following what they find genuinely interesting, and it does not survive being assigned.

The cost is that we cannot take everything that comes in. That is the trade, and we would make it again.

The hard part was never the model. It is everything the model touches.

Which is why we sell the integration, the evaluation and the handover, not the demo.
Where we are And how that works

Based in Indore, Madhya Pradesh.

Vibencode Pvt. Ltd., registered at Sagar, Madhya Pradesh. The team is spread across India and works remotely, from wherever the engagement needs it.

Remote by default, on site when it earns its cost.

Discovery and handover are the two phases where being in the room genuinely changes the outcome. The rest rarely needs a flight.

You will always know where it is.

A shared channel and a written update every week. Most of what people dislike about consultants is not the work, it is not knowing.

Bring us the part you can only describe.

A team to work with. Not a platform to subscribe to.