Vote for your favorite SkillMD. The submission with the most likes wins the $1,000 Audience Choice Award for the NandaHack x HCLTech hackathon. Voting is open through September 25.Vote now →
Nanda Town

Agentic Commerce Hackathon

Agentic Commerce Hackathon — July 31 – August 2, 2026, Online, 48-hour event. Build AI agents that can discover, decide, and complete transactions.

July 31 – August 2, 2026 · Online, 48-hour event

Apply on Devfolio →How to submit →

Prize pool

$73,500+

All approved participants also receive $100 in OpenAI credits.

Agentic Commerce Hackathon

$10,000

OpenAI

$39,000

Visa

$5,000

Linq

$6,000

Localhost

$5,000

Project NANDA

$1,000

Senso

$7,500

Project NANDA track · $1,000

Best Prava Adapter for NANDA Town

The primary challenge is to build a reliable and reusable Prava payments adapter for NANDA Town, enabling agents to quote, pay, verify transactions, and handle failures using Prava's Agentic Payments Sandbox.

Teams can go further by using their adapter to run commerce simulations involving autonomous buyers, sellers, negotiations, budgets, retries, and other real-world payment conditions.

Submissions are evaluated on

  • Quality and reliability of the adapter
  • Security, authorization, and failure handling
  • Ease of installation and reuse
  • Successful Prava sandbox transactions
  • Quality of any agentic-commerce scenarios or simulations built on top

The strongest submission should make it easy for future NANDA Town builders to use Prava as a payment layer.

How to submit

Both parts are required for this track: the pull request and the Devfolio submission.

  1. Build the adapter as a Nanda Town payments-layer plugin. The payments layer covers quote, pay, verify, and refund. Follow the Writing a plugin guide in the docs.
  2. Connect it to Prava’s Agentic Payments Sandbox. Show at least one successful sandbox transaction in a scenario run. Include the scenario and a test, and handle at least one failure case.
  3. Open a pull request to github.com/projnanda/nandatown. It should carry the plugin, the scenario and test, and a README that explains installation and reuse. Your pull request shows up in the Payments layer of the PR gallery.
  4. Submit your project on Devfolio as your official hackathon entry. Link your pull request in the Devfolio submission so the judges can find the code.

What to build

Agents that transact.

Participants create working products where agents can "shop across merchants, book a trip, manage a subscription, procure software, reserve a table or create a new kind of commerce experience."

Prava is a payments and trust layer that lets AI agents make purchases, available via API/SDK or MCP & CLI.

Eligibility

Open to students, developers, designers, operators, researchers, and founders. Build solo or in teams of up to 4 — every team member applies and is accepted individually.

  1. 01Apply via Devfolio (rolling review basis).
  2. 02RSVP within 3 days if accepted.
  3. 03Join the Discord — membership is mandatory.
  4. 04Build meaningful new work during the hackathon window.
  5. 05Demonstrate agent-enabled transaction completion using Prava.

Judges & speakers

Prof. Ramesh Raskar

MIT Media Labs, Project NANDA founder

Justin Leung

Visa, Agentic Commerce Principal

Josh Flayhart

Linq, Staff Engineer

Ujwal Chaudhari

Visa, Product

agentic-commerce.devfolio.co

Apply on Devfolio →