The next generation of AI applications will need more than static knowledge. They will need to understand what is happening now, respond to continuously changing information, and take action as events unfold.
At Data Streaming Summit, developers, data engineers, and AI builders will have an opportunity to explore this future firsthand during a full-day, in-person hackathon focused on building agents powered by real-time data.
One day. One room. Real-time data. Agents built to act.
Whether you arrive with a specific idea or simply want to experiment with streaming data and AI, the hackathon will provide the environment, technology, guidance, and support you need to turn an idea into a working prototype.
From Streaming Events to Intelligent Action
Most traditional AI applications rely on information that was collected and indexed in the past. But many important decisions depend on what is happening right now:
- A customer's behavior suddenly changes.
- A payment pattern becomes unusual.
- A device begins reporting abnormal readings.
- An infrastructure service starts to degrade.
- Inventory drops unexpectedly.
- A new security threat emerges.
- A credit score or account-risk indicator changes.
Real-time agents can continuously monitor signals like these, investigate the surrounding context, recommend a response, and help people take action.
During the Data Streaming Summit Hackathon, participants will combine live Kafka data, continuous SQL processing, and agentic AI to build applications that understand events as they happen.
Start with a Ready-to-Build Environment
Every team will receive access to a dedicated StreamNative environment that includes:
- A Kafka cluster
- Pre-created Kafka topics
- Continuously generated streaming data
- SQL Workspace
- Agent Workspace
- StreamNative Managed Agents
- Starter examples and technical guidance
Instead of spending the day configuring infrastructure, participants can begin by exploring live streams, creating continuous transformations, and building agents.
The provided datasets will include realistic, interconnected events that can be queried and correlated across multiple topics. Participants can use these streams as the foundation for their projects, extend the supplied scenarios, or introduce data for an entirely different idea.
Learn How Real-Time Agents Work
The day begins with an introduction to StreamNative Managed Agents and the role of real-time data in agentic applications.
StreamNative experts will demonstrate how an agent can use a continuously changing streaming context to:
- Monitor live activity
- Detect meaningful changes
- Investigate related events
- Explain what happened
- Recommend an appropriate response
- Request human approval before taking consequential action
Participants will also learn about the environment, available development tools, and the workflow they will use throughout the day.
Build Your First Agent
Following the introductory session, participants will complete a guided hands-on tutorial.
The tutorial will walk through the complete development flow:
- Explore live events across Kafka topics.
- Query streaming data using SQL Workspace.
- Correlate information from multiple streams.
- Create a materialized view that updates continuously.
- Turn the materialized view into a rich context for an agent.
- Configure and test a Managed Agent.
- Ask the agent to monitor, investigate, and explain changing conditions.
- Introduce a human-approval step before an action is completed.
The tutorial will provide a working foundation that teams can extend during the hackathon.
Bring an Idea---or Discover One During the Day
After the tutorial, participants will form teams and begin building.
There is no prescribed challenge or required industry scenario. Teams can build any agent that applies real-time data to a meaningful, real-world problem.
You might explore an idea related to financial services, security, commerce, infrastructure, customer experience, developer productivity, IoT, operations---or something entirely different.
Participants can use StreamNative's Agentic AI capabilities or connect StreamNative's real-time data platform to a third-party agentic AI platform. The challenge and technology path are yours to define.
The goal is not to build a production-complete application in a few hours. The goal is to demonstrate a compelling idea through a functional prototype:
- What real-time signals does the agent monitor?
- How does continuous SQL prepare the agent's context?
- What change or condition does the agent recognize?
- How does it investigate and explain the situation?
- What decision, recommendation, or action does it enable?
Build with Experts in the Room
StreamNative product and engineering experts will be available throughout the day to help teams with:
- Kafka and streaming-data architecture
- Topic and schema design
- SQL Workspace
- Continuous queries and materialized views
- Agent Workspace
- Managed Agent configuration
- Third-party agent integrations
- Application development
- Troubleshooting and debugging
Each team will have access to the platform resources and technical support needed to keep moving from idea to demonstration.
Show What You Built
The day concludes with team demonstrations.
Each team will take the stage to present:
- The real-world problem it chose to address
- The streaming data used by the solution
- How the agent understands changing conditions
- How the application investigates, recommends, or acts
- What the team would build next with more time
The strongest demonstrations will not necessarily be the most complex. A focused solution that clearly shows why live data matters can be just as powerful as an ambitious technical implementation.
Built for Every Level of Experience
You do not need to be an AI expert to participate.
The morning sessions will provide a practical foundation, and StreamNative engineers will be available throughout the event. Experienced builders can go deeper into custom processing and integrations, while participants who are newer to streaming data or agents can begin with the guided project and starter resources.
You can attend with an existing team or join as an individual. Individual participants will have an opportunity to connect with other builders and form teams at the beginning of the hackathon.
Recommended team size: 2 participants.
How Projects Will Be Judged
A panel of judges will evaluate each project based on:
- Real-time data usage: How effectively the solution uses continuously changing streaming data
- Agent capabilities: How well the agent monitors, investigates, explains, recommends, or acts
- Usefulness and impact: The relevance of the problem and the value of the proposed solution
- Creativity and innovation: The originality of the idea and overall approach
- Technical execution: How effectively the team brings together streaming data, processing, and agentic AI
- Demonstration quality: How clearly the team presents the problem, solution, and working prototype
Judges will consider the complete outcome---from the strength of the idea to how effectively it is demonstrated.
Build, Compete, and Win
The Hackathon isn't just a chance to experiment with real-time agents---it's also an opportunity to compete, win, and continue the experience at Data Streaming Summit.
🏆 Compete for prizes
Bring your best idea and build. Based on the final judging scores, first-, second-, and third-place winners will be selected, with some seriously exciting surprise prizes up for grabs.
We're keeping the details under wraps for now, but you'll want to compete for these. Prizes will be revealed during the Data Streaming Summit keynote.
🎟️ Earn free admission to Day 2
All participants who submit their work during the Hackathon will receive free admission to the Data Streaming Summit Conference on Day 2 to keep learning, connect with the community, and hear from leaders shaping the future of data streaming and agent infrastructure.
How the Day Works
- 9:00--10:00 AM --- Learn: Introduction to Managed Agents, the development environment, and an end-to-end demonstration.
- 10:00--11:00 AM --- Get Hands-On: Guided tutorial using Kafka, SQL Workspace, and Managed Agents.
- 11:00--11:15 AM --- Break: Connect with other participants and prepare for the build session.
- 11:15 AM--4:00 PM --- Build: Form a team, select a problem, and create a working prototype with support from StreamNative experts.
- 4:00--5:00 PM --- Demonstrate: Present your project and show what becomes possible when agents understand events as they happen.
What to Bring
Bring:
- Your laptop and charger
- Your preferred development tools
- An idea you want to explore---or the curiosity to discover one
- A willingness to build, experiment, and collaborate
StreamNative will provide the streaming environment, sample data, workspaces, starter resources, and onsite technical support.
Build What Comes Next
Real-time data gives agents the context to understand what is happening now. Continuous processing helps them recognize what has changed. Agentic AI turns that understanding into investigation, recommendations, and action.
At the Data Streaming Summit Hackathon, you will have one day to bring those capabilities together and create something real.
The event is in person and limited to 50 participants. Reserve your place and come build the future of real-time agents.
