INDEPENDENT PRODUCT BUILDER · AI · 0→1 INDIA / 2026

I AM A BUILDER.

I help founders turn ambitious ideas into shipped products.

I spent two years building startups of my own. Now I use that product judgment, engineering depth and founder mindset to help ambitious teams go from idea → working product.

6 products shipped · 2 years building full-time · Paid startup consulting
Work with me See my work
SCROLL TO EXPLORE ↓
01 / WORK WITH ME

I help founders build what's next.

The best projects usually start with something messy: an idea, a prototype, a product that's almost working, or a technical problem nobody owns yet.

That's where I'm useful.

BEST FIT

Small teams.
Ambitious problems.

I work best with founders and small teams moving quickly on products where the requirements are still evolving—especially early-stage and AI startups, or teams turning a prototype into a real product.

If the problem is ambiguous and ownership matters more than a perfectly written spec, we'll probably work well together.Book a 30-minute call
02 / SELECTED WORK

Don't take my word for it.
Here's what I've built.

Four ambitious products taken from a loose question to working software across AI, mobile, web, and operating systems.

01 / 2024SHIPPED

LLM CHATBOT PLATFORM

Watch full demo

SmartChat

Personalized AI chatbots, grounded in your knowledge.

A web + mobile platform for creating AI chatbots with user-uploadable knowledge.

smartchat.app
How can I help?Ask anything about your documents.
LLM CHATBOT PLATFORMSmartChat
THE PROBLEM
Useful AI assistants need an answer source people can trust—not another empty chat box.
WHAT I BUILT
Independently scoped and shipped the web and mobile product with production authentication, Dockerized services, and a RAG pipeline using MongoDB Vector Search and OpenAI embeddings.

Shipped across web and mobile, used by real users, and improved through direct product feedback.

TECHNICAL DEPTHCode

Production-ready Sign in with Apple and Google, containerized deployment, vector search, and cross-platform delivery—built end to end.

ReactReact NativeNode.jsDockerMongoDB Vector SearchOpenAIAWSApple + Google Auth
02 / 2025SHIPPED

AI RESUME BUILDER

JobOS

Prompt to resume.

An AI resume product spanning a mobile app and companion browser extension.

jobos.app
KAPIL VERMA
Optimize this resume for
Product Engineer →
AI RESUME BUILDERJobOS
THE PROBLEM
Tailoring a strong resume for each role is tedious, repetitive work with meaningful consequences.
WHAT I BUILT
Built an end-to-end workflow from PDF ingestion and structured content to LaTeX generation and LLM optimization, plus a companion browser extension for capturing job context.

Shipped as an iOS app with a companion browser extension, then refined through feedback from real users.

TECHNICAL DEPTHCode

A multi-stage document pipeline combining PDF parsing, structured generation, LaTeX compilation, LLM optimization, and a companion browser extension.

React NativeLLMsPDF ParsingLaTeXBrowser ExtensioniOS App Store
03 / 2026SHIPPED

VOICE NOTES → MINDMAPS

Watch full demo

Yapwise

Talk it out. See the structure.

An iOS app turning spoken notes into mind maps, action items and reminders.

yapwise.app
Voice note00:43
project ideas
VOICE NOTES → MINDMAPSYapwise
THE PROBLEM
Loose thoughts are easy to capture and surprisingly hard to turn into a next step.
WHAT I BUILT
Built a voice-first iOS experience that transcribed speech, generated structured notes, visualized mind maps, and extracted action items and reminders.

Shipped to the iOS App Store, tested with real users, and refined through direct product feedback.

TECHNICAL DEPTHCode

A native voice pipeline connecting speech recognition, LLM-powered structured generation, interactive mind-map visualization, and action extraction.

SwiftSpeech RecognitionLLMsStructured GenerationMind MapsiOS
04 / 2026PROTOTYPE

AI-NATIVE OPERATING SYSTEM

Watch full demo

DragonOS

An OS for LLMs.

An exploration of what happens when AI agents are native to the operating system—not bolted onto applications.

dragonos.app
dragon@os:~$ Build a brief for tomorrow's meeting_

Opening browser · drafting document · done
AI
AI-NATIVE OPERATING SYSTEMDragonOS
THE PROBLEM
Traditional software makes people translate intent into clicks. Agents should be able to work directly from intent.
WHAT I BUILT
Built a working Linux-agent prototype capable of writing documents, interacting with the operating system, and browsing the web directly from a prompt.

Took an AI-native operating-system concept from pitch to a working prototype.

TECHNICAL DEPTH

System-level agent orchestration across local and cloud models, browser automation, document generation, and native Linux APIs.

LinuxAgent OrchestrationLocal + Cloud LLMsBrowser AutomationSystem APIsDocument Generation

THE GRAVEYARD

Things I killed.

Not every idea deserves another feature.

One of the most useful things I learned during two years of building was when to stop.

  • Vrowser— archived
  • Bro— archived
  • Career Explorer— archived
  • Other early experiments— archived
03 / THE LONG WAY ROUND

702 days.

I skipped placements to see how far I could take my own ideas.

  1. 2024

    Graduated from IIT Hyderabad

  2. 01

    Skipped campus placements

  3. 02

    Started building full-time

  4. 03

    6 products shipped

  5. 04

    Reached real users and gathered direct product feedback

  6. 05

    2 apps shipped to the iOS App Store

  7. 06

    Built an AI-native Linux agent

  8. 07

    Some ideas worked. Several didn’t.

  9. 2026

    Returned with two years of scars, code, product judgment and a much better understanding of what people actually use.

YOUTUBE STORY / 2026I chased my dream
for 702 days.0 → 1
Watch the full story

Those 702 days taught me more than how to build software. They taught me what to build, what not to build, and how quickly assumptions fall apart once real users arrive.

Now I bring those lessons into the products I build with other founders.
04 / SELECTED CLIENT WORK

Trusted with
the messy part.

Brave was building a 2-in-1 computing device and needed someone who could move across infrastructure, product decisions and implementation.

BRAVE.TECH

Technical + Product Consulting

2026

I worked directly with the team across the software stack, taking ambiguous requirements through to working systems and a functional product prototype.

  • Owned work across AOSP, AWS bare-metal infrastructure, Cuttlefish virtual devices and custom ROMs.
  • Defined, designed and built the device’s pre-installed note-taking experience from requirements to a functional prototype.
  • The work resulted in offers to continue full-time across both engineering and product.
Discuss a project
EDUCATION

Indian Institute of Technology Hyderabad

B.Tech2020 — 2024

05 / CAPABILITIES

What I can own.

You can bring me in before the requirements document exists. I can help define the product, make the technical calls and carry it through to launch.

Product

From vague idea → clear MVP.
  • Product discovery
  • Scoping and user flows
  • Prioritization
  • Go-to-market thinking

AI

From “we should add AI” → useful interaction.
  • LLM product strategy
  • RAG and embeddings
  • Agent workflows
  • AI-native interfaces

Engineering

From prototype → working software.
  • Web and mobile
  • Frontend and backend
  • Cloud and deployment
  • System-level products

Execution

From requirements → something shipped.
  • Customer conversations
  • Cross-functional delivery
  • App Store publishing
  • Iteration after launch
06 / NOTES

Thinking in public.

I'm interested in AI as a new interface for software—not just a feature to put behind an API call.

Elsewhere:Medium X

07 / ABOUT

The path was
not linear.

I'm Kapil Verma. I studied Chemical Engineering at IIT Hyderabad and then spent two years doing something almost completely unrelated to my degree: trying to build companies.

I skipped placements, built six products, shipped apps, talked to users, pitched investors, killed ideas and learned the difference between software that works and products people actually want.

Along the way, a startup began paying me to solve its product and engineering problems too. Today I work with founders on ambitious 0→1 products, especially around AI.

Bring me the problem before the spec exists.
IIT HyderabadAWS for StartupsMongoDB for StartupsNotion for StartupsShipped and maintained productsUCMAS Mental Arithmetic · State Champion / National Runner-up
08 / CONTACT

Have something
worth building?

If you're a founder working on an ambitious product and need someone who can think through it and build it, let's talk.

Book a 30-minute call

I typically work directly with founders and small product teams.

LinkedIn GitHub YouTube Instagram