Building an AI application is easier than ever. Building one that works in the real world is still a challenge.
Developers today have access to frontier models, increasingly capable AI coding assistants, and cloud infrastructure that can scale on demand. But moving from an impressive prototype to a reliable application requires solving a much harder problem: wrangling all the pieces.
Where does the data live? How does it flow into applications? How do models access the information they need? How do developers combine APIs, agents, event processing, and application logic into a system that can respond quickly and operate reliably?
The challenge is no longer just building a model. It is building the complete AI system around the model.
At VAST, we believe developers should spend more time building and less time stitching together disconnected infrastructure. That belief is at the core of the VAST AI OS: a unified platform designed to bring together data, compute, database capabilities, and AI services so developers can build and deploy modern AI applications faster.
That is why we are launching the VAST Builders Challenge: a three-city, hands-on series for AI engineers, software developers, data scientists, and platform architects who want to build real applications using a modern AI stack.
This is not a conference where you watch demos and listen to talks. It is a day dedicated to building, experimenting, and collaborating with developers solving similar challenges.
Building with the Modern AI Stack
Today's AI applications bring together many different technologies: data platforms, accelerated computing, AI models, developer tools, and application frameworks.
The VAST Builders Challenge gives developers the opportunity to experience how these components work together.
Participants will build applications using the VAST AI OS and technologies from our ecosystem partners, including:
VAST Data: A unified AI data platform that provides the foundation for managing, processing, and activating data for AI workloads
CoreWeave: the Essential AI cloud for hosting models - both self-managed and serverless - including Weights & Biases AI developer tools for experiment tracking, model evaluation, auto research, and collaborative ML development
Cursor: AI coding agent that helps developers write, review, and ship software faster
NVIDIA: Accelerated computing, AI software tools and open models—such as Metropolis VSS Blueprint, Cosmos, Nemotron and NemoClaw, that enables developers to build and deploy advanced AI applications.
Learn. Build. Demonstrate.
The Builders Challenge event follows a simple idea: the best way to understand modern AI systems is to build with them.
Participants will learn how modern AI applications are architected, how data flows through these systems, and how different components come together to create production-ready systems.
Whether you're already building AI applications or just beginning to explore modern AI infrastructure, the workshop is designed to be approachable while still offering meaningful technical depth. Experience with LLMs, embeddings, or event-driven systems is helpful, but not required. We'll provide datasets and sample content to help you get started, and if you have your own video clips or data you'd like to build with, you're welcome to bring those as well.
Developers will work in teams to create AI applications using the tools and technologies introduced during the workshop. Projects may include agentic applications, RAG workflows, data-intensive applications, multimodal experiences, or other AI-powered solutions.
Throughout the day, experts from VAST, CoreWeave, Cursor, and NVIDIA will be available to answer questions, share best practices, and help teams overcome technical challenges.
At the end of each event, participants will demonstrate their applications to a panel of judges and the broader builder community.
A Community for AI Builders
The questions developers face today are increasingly complex:
How should AI applications access enterprise data?
How should organizations manage the growing number of AI workloads?
How do teams move from experimentation to production at scale?
The VAST Builders Challenge is designed to create that environment, a place where builders can exchange ideas, experiment with new approaches, and see what becomes possible on modern AI infrastructure.
Build. Demo. Compete.
Each city will conclude with a project showcase and competition.
Teams will be evaluated on creativity, technical implementation, and the real-world impact of their applications.
First-place winners in each city will receive an NVIDIA DGX Spark(™) to continue building after the event, with additional prizes to be announced.
More importantly, every participant will leave with hands-on experience building AI applications, new connections with fellow developers, and a deeper understanding of the architecture patterns shaping the future of AI.
Join the VAST Builders Challenge
The inaugural VAST Builders Challenge will take place this October in three cities:
San Francisco - October 2
New York City - October 9
London - October 17
Space is limited to keep each workshop interactive and hands-on. If you are building AI applications, exploring new AI architectures, or looking to better understand how data and AI infrastructure come together, we invite you to join us.
I believe the future of AI will be built by communities of developers who understand how data, compute, and AI models work together. That is what the VAST Builders Challenge is about.
Bring your ideas. Bring your curiosity. Let's build!
