<aside> <img src="/icons/location_green.svg" alt="/icons/location_green.svg" width="40px" /> Antler VC

</aside>

<aside> <img src="/icons/calendar_green.svg" alt="/icons/calendar_green.svg" width="40px" /> April 10 - 12

</aside>

<aside> <img src="/icons/profile_green.svg" alt="/icons/profile_green.svg" width="40px" /> In Person

</aside>

Welcome to the Spark Hack Series New York - Hosted by NVIDIA, Acer & Antler! We’re excited to host a diverse group of builders for a weekend.

Table of Contents

Hacker Resources

Demo Video Instructions

DGX Spark Playbooks (1)

DGX Spark Livestreams (1)

OpenClaw Resources

<aside> <img src="/icons/headset_green.svg" alt="/icons/headset_green.svg" width="40px" />

Join The Discord!

This will be the easiest way to communicate with our team, get updates on the hackathon, and connect with other hackers. Please join the #spark-hack-nyc channel ASAP!

</aside>

<aside> <img src="/icons/alien-pixel_green.svg" alt="/icons/alien-pixel_green.svg" width="40px" />

SUBMIT YOUR PROJECT!

</aside>

<aside> <img src="/icons/map_green.svg" alt="/icons/map_green.svg" width="40px" />

Getting Situated

Wifi

</aside>

<aside> <img src="/icons/list_green.svg" alt="/icons/list_green.svg" width="40px" />

Build Challenges

Context

All submissions must align with one of the three Challenge Tracks below. Leveraging open data from the City of New York, these tracks guide submissions across diverse facets of urban life and community growth.

Each track defines a theme of impact, not the scope of your idea - teams are free to build any solution using any of the open datasets linked above.

1. Human Impact

Focus: Enhancing the health, safety, and economic well-being of New York residents.

• The Goal: Use data to build tools that improve individual lives, streamline access to social services, or promote economic equity.

2. Environmental Impact

Focus: Driving sustainability, urban resilience, and smarter resource management.

• The Goal: Address the "green" future of the city by optimizing how we use energy, manage waste, and move through urban spaces.

3. Cultural Impact

Focus: Celebrating the soul of New York through arts, recreation, and community identity.

• The Goal: Use data to make the city’s rich cultural and recreational offerings more accessible, or to preserve the history of its unique neighborhoods.

Bounties

Most Impactful Use of OpenClaw - Build a new skill for the OpenClaw Skills Hub that powers a real-world impact application. We're looking for productivity and application-specific skills that help agents do meaningful work — think coordinating community meetups, managing volunteer recruiting, or supporting environmental initiatives. You'll be running OpenClaw on the Acer Veriton GN100 via NVIDIA NemoClaw for easy setup and security.Winner gets a RTX 5090!

Most Likely to Be a Unicorn - Build a project that has legitimate potential to turn into a Unicorn business. Winner gets dinner with the Antler team and automatic acceptance to the Founder in Residence program

Least Likely to get Hacked - Run Pensar on your build and fix the issues it surfaces. $500 cash and $500 Pensar credits for the winning team.

</aside>

<aside> <img src="/icons/judicial-scales_green.svg" alt="/icons/judicial-scales_green.svg" width="40px" />

Judging Criteria

Philosophy

We are judging Systems Engineering. A winning project isn't just a slide deck or a simple API wrapper; it is a functioning system that ingests raw data, processes it locally using the Acer Veriton GN100 AI Mini Workstation, and produces a valuable result.

The Scoring Breakdown (100 Points Total)

1. Technical Execution & Completeness (30 Points)

Did they actually build a working, complex system?

• 15 pts - Completeness: Does the system successfully complete the full data workflow without crashing?
• 15 pts - Technical Depth: Is there significant engineering "under the hood"? Did they build a complex pipeline (e.g., Simulation, RAG, Fine-Tuning, or Custom Logic) rather than just a simple static dashboard or basic API wrapper?

2. NVIDIA Ecosystem & Spark Utility (30 Points)

Did they leverage the unique hardware and software provided?

• 15 pts - The Stack: Did they use at least one major NVIDIA library/tool? (e.g., NIMs, RAPIDS, cuOpt, Modulus, NeMo Models). Note: Merely calling GPT-4 via API gets 0 points here.
• 15 pts - The "Spark Story": Can they articulate why this runs better on a DGX Spark?
  • Examples: "We used the 128GB Unified Memory to hold the video buffer and the LLM context simultaneously" or "We ran inference locally to ensure privacy/latency."

3. Value & Impact (20 Points)

Is the solution actually useful?

• 10 pts - Insight Quality: Is the insight non-obvious and valuable? (e.g., "Traffic jams happen at 5 PM" is obvious. "Rain causes specific stalls on this specific ramp" is valuable).
• 10 pts - Usability: Could a real City Planner, or Factory Foreman actually use this tool to make a decision tomorrow?

4. The "Frontier" Factor (20 Points)

Did they push the boundaries?

• 10 pts - Creativity: Did they combine data or models in a novel way? (e.g., Using vision models to "read" traffic maps).
• 10 pts - Performance: Did they optimize the system for speed or scale? (e.g., "We optimized the simulation to run at 50x real-time speed"). </aside>

Agenda - Day 1