About the challenge
First Visions: Danta is our first AI hackathon, built for devs who want to create real, working AI agents on Maestro.
This September we’re bringing participants from LATAM and SEA together to build solutions in Finance, Education, and Productivity that are ready for others to use.
-
Dates: Sept 19-21 (build) • Sept 22-25 (evaluation & public voting) • Sept 26 (winners)
-
Format: Hybrid. Work in teams of 2-3. No solo entries.
-
Prizes:
1st Place - $1,000 + 300 Maestro Credits
2nd Place - $500 + 200 Maestro Credits
3rd Place - 100 Maestro Credits
All participants receive an Agent Builder Certification. -
Judging focus: Engineering strength, market readiness, and clear documentation.
Get started
Planning is at the essence for this Hackathon, better get a headsup first and plan accordingly to have an advantage over the other teams. Here are some tips for you:
- Get Maestro Access: Sign up at hack.dantalabs.com to claim your API keys and credits. Get Tokens
- Join our Community: Hop on our Discord for announcements, team formation, and support. Join Discord Channel
- Read the Docs: Follow the Get Started section, deploy your first “Hello World” agent to familiarized yourself, and check the examples repo. Read Docs
-
Set Up Early: Install Maestro SDK (pip install dantalabs), configure keys, and test deploy before the start.
In-Person Experience
We’re hosting an in-person hackathon session in Ho Chi Minh City for the full 3 days. If you want to join us on-site, you’ll need to register through the local event page. Spots are limited, so make sure to secure yours early!
Requirements
What to Build
Using Maestro, we want to see projects that take full advantage of the platform’s building and hosting capabilities. This can be something you’ve been meaning to publish but never had the time for, or a fresh idea you want to bring to life quickly.
Here are some categories to inspire you:
Productivity - Create tools that make daily tasks faster and smoother for a specific type of user. This could be automations for schoolwork, small businesses, or niche workflows. Think email automation, task coordination, CRM updates, or social media integrations.
Finance - Build something that improves personal or corporate finances. From a simple expense tracker to a complex credit assessment agent, if it helps users save or earn money, it’s in scope.
Education - Build tools that help people learn, stay informed, and keep up with the latest in AI and beyond. This could mean AI explainer bots, interactive study aids, or research companions.
Wildcard - Got an idea that doesn’t fit the above categories? Go for it. As long as it uses Maestro and delivers value to its users, it’s welcome here.
What to Submit
We want to see your project in action and understand exactly how you built it. Make it easy for the judges (and future users) to test, learn from, and build on your work. Here’s what your submission should include:
- Devpost Project Page - Give your project a clear name, a shortdescription, and explain the problem it solves. Focus on what it does, why it matters and who it’s for.
-
Public GitHub Repository - Your repo must be public with an approved OSI Open Source License so others can review and learn from it. The README should cover:
-
Setup Instructions: How to install and run your agent.
-
Usage Guide: How a user or developer can interact with it.
-
Example API Calls: Sample requests and responses so others can integrate it quickly.
-
- Demo Video (2–3 minutes) – Show your agent in action from start to finish. Walk us through what it does, how it works, and why it’s useful. Keep it concise and focused. Upload to YouTube as Public or Non-Listed. Private will not be accepted.
- Live Endpoint or Agent ID – Share a working Maestro endpoint or ID so we can test your deployment directly.
Prizes
First Prize - $1.000 USD + 300 Maestro Creds
We will also provide 300 credits for Maestro to keep on building!
Second Prize - $500 USD + 200 Maestro Creds
Plus 200 credits for Maestro for your projects to keep on rolling!
Third Prize - 100 Maestro Creds
100 Credits on Maestro to explore the platform and create more Agents
Devpost Achievements
Submitting to this hackathon could earn you:
Judges
Santiago Canchila
CTO - Danta Labs
Krishna Chaitanya
Senior Software Engineer - Meta
Modaldo Tuñon
PhD - Tecnological University of Panama
Ashwin Shankar
Sr. Data Science Manager - Atlassian
Ali Rathore
Lakeshore Labs
Vivek Banerjee
Ph.D., h.c. – SAP Enterprise Integration Architect
Juan Carlos Correa
PhD - Tecnológico de Monterrey
Ngoc Doan
Software Engineer at DoorDash
Alexander Farr
Co-Founder & Agentic Outreach - Clera
Daniel Wintermeyer
CTO - Clera
Judging Criteria
-
Problem/Solution Fit - 25%
How well the agent solves a clearly defined problem for its intended audience. -
# of Agents in the System - 10%
Total distinct agents in the solution and how they interact. -
# of Connected External APIs - 10%
Total unique, functional integrations with external services. -
User Usability - 15%
Ease of use, onboarding flow, and clarity of instructions for end-users. -
Agent Capacity - 20%
The percentage of the process handled independently by the agents without human intervention. -
Public Vote of Confidence - 20%
As a community, it is chosen which of the agents is the most interesting use case
Questions? Email the hackathon manager
Tell your friends
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.