Menu
  • AI-native founder
  • Product architect
  • Hyderabad, India

Varun ChennuriI build production AI products in days, with an AI engineering team I designed and run.

Founder of XusCorp. Claude Code and Codex build in parallel and review each other's code, and I take the risky calls. Hiring an AI product developer in India? Start with a 7-day sprint.

Start a 7-day sprintRead how I build
[PLACEHOLDER: SKETCHED PORTRAIT]
Operator_ID: 001Hyderabad
Builder

Now building

Karkaana, Tinder for factories in India.

86 commits in about 17 days. Pre-launch.

Proof, not claims

The numbers so far

Every number comes from the build it describes, and each card links to where it came from.

Case files

Featured builds

A factory marketplace, a Shorts channel that makes itself, and voice agents that speak 12+ Indian languages.
All case files

  • Marketplace · India

    Pre-launch

    Karkaana

    Tinder for factories in India. Type what you want to make, swipe through real manufacturers, send a quote request.

    94,000+real MCA-registered factories indexed

    • Marketplaces
    • Manufacturing
    • India

  • Media AI · Automation

    Live

    Why Pip?

    A curious-facts Shorts channel run by an automated AI pipeline, from trending topic to daily upload.

    69Shorts produced since late October 2026

    • Media AI
    • Automation
    • YouTube Shorts

  • Voice AI · India

    Live pilot

    Zaya

    Real-time voice agents in 12+ Indian languages that handle bookings and negotiation.

    12+Indian languages

    • Voice AI
    • 12+ languages
    • Real time

The method

An AI team, not a chatbot

Claude Code and OpenAI Codex each build a different part of a task in parallel, then review each other's code. Two different models catch each other's mistakes. Then tests, then a pull request.
Read how I build

  1. Build in parallel

    Claude Code and OpenAI Codex each build a different part of the task at the same time.

  2. Cross-review

    Each model reviews the other's code. Two different models catch each other's mistakes.

  3. Test

    Before any bug gets fixed, a failing test has to prove it exists. Then the tests run.

  4. Human gate

    I step in only for the risky calls: data migrations, deploys, money, and anything ambiguous.

Every task gets a line budget. I added that rule after one AI-built project grew past 16,000 lines.

Contact

Let's build it this week

Email or WhatsApp works. Or book a sprint call and bring the one promise your product has to keep.

Reach me directly

Next step

Got an AI product to ship?

Seven days from framing to delivery. A two-model AI engineering team does the building, and I sign off on every risky call.

Start a 7-day sprintRead how I build