Next Hackathon
Band of Agents Hackathon
Starts Jun 12, 2026
Transforming Enterprise Through AI
Build AI Agents for Security, Edge & Robotics
π May 11β19, 2026
β’ May 11 β 19: Online Build Phase.
β’ May 18 (Hybrid Build Day β Online & Onsite)
β’ May 19 (Demos & Awards β Online & Onsite)
π Hybrid event. Join online from anywhere or build onsite at the venue.
π Winners will be announced live on stage at the AI & Big Data Expo North America on May 19
π On-site Venue (May 18β19):
San Jose McEnery Convention Center, CA, USA
π€ Join solo or with a team.
π FREE AI & Big Data Expo North America ticket included for participants that are able to join the on-site phase.
πCompete for a $10,000 prize pool
π§βπ» Apply now to design, prototype, and showcase enterprise-ready innovation.
Hackathon Overview
Our AI hackathon brought together a diverse group of participants, who collaborated to develop a variety of impressive projects based on:
3124
Participants
893
Teams
291
AI Applications
Winners and Finalists
- We are in the process of selecting the finalist teams.
- Your voice matters! Vote on your favorite projects in the section below.
- Join us for the winner announcement stream, which will be streamed live on Twitch.
This event has now ended, but you can still register for upcoming events on lablab.ai. We look forward to seeing you at the next one!
Check out Upcoming Events β
AI and Big Data Expo North America is a leading event focused on enterprise AI deployment, taking place as part of TechEx North America. The event explores topics ranging from large language models (LLMs) to Physical AI and scaling real-world AI systems in production environments.
π FREE AI & Big Data Expo North America ticket included for participants that are able to join the on-site phase.
Join Fortune 500 leaders, AI innovators, and global technology partners as they move artificial intelligence from pilot projects to operational reality. The agenda covers generative AI, autonomous systems, AI governance, and enterprise infrastructure.
The event is co-located with the Intelligent Automation and Robotics Expo, with content spanning cognitive systems, robotics, and autonomous technologies. Attendees will explore how global enterprises are redesigning operations in the AI era.
With 8,000+ attendees, 200+ exhibitors, and 250+ expert speakers from global organisations such as Google, NASA, and Microsoft, AI and Big Data Expo is designed for practical discussions around enterprise AI deployment and scalable innovation.
Winners will be announced live on stage at the AI & Big Data Expo North America on May 19
Media & Community partners
Build the Next Intelligent Enterprise Solution
Design and prototype a scalable solution that leverages AI agents, data, or intelligent systems to solve real-world enterprise problems.
Your project should demonstrate practical value, strong technical thinking, and a working demo.
Tracks
Select one primary track:
π Track 1: Agent Security & AI Governance
Build AI systems that enterprise security teams can actually trust and deploy.
Powered by: Veea
Focus areas:
β’ Guardrails and safety layers for agentic workflows
β’ Monitoring and observability tools for AI agents (detect hallucination, drift, misuse)
β’ Access control and permission frameworks for multi-agent systems
β’ Audit trails and explainability tooling for regulated industries (finance, healthcare, legal)
β’ Red-teaming frameworks for testing agent robustness against adversarial inputs.
π€ Track 2: AI Agents with Google AI Studio
Build production-ready agent workflows using Gemini models.
Powered by: Google DeepMind & Google AI Studio
Focus areas:
β’ Multi-agent systems using Gemini
β’ Long-context document processing (contracts, reports)
β’ Code generation and developer workflow agents
β’ Internal AI tools (knowledge bases, ops automation)
β’ Enterprise integrations (CRM, ERP, APIs)
Build intelligent systems that interact with or simulate the physical world.
Focus areas:
β’ AI-powered robotics control systems
β’ Simulation environments for training/testing
β’ Digital twins for industrial environments
β’ Vision-language models for real-world tasks
β’ Human-robot collaboration interfaces
π Track 4: Data & Intelligence
Turn enterprise data into actionable intelligence.
Focus areas:
β’ RAG systems over proprietary or multi-source data
β’ AI-powered data pipelines and validation
β’ Analytics agents for natural language querying
β’ Anomaly detection and forecasting
β’ Knowledge graph extraction from documents
Your mission is to build and launch an AI-native product using X402 Payments - a programmable payments infrastructure designed for the next generation of agentic and automated economies.
In this edition of Launch & Fund Your Startup, youβll have 6 days to validate an idea, build a working prototype, and present a credible path toward real users and on-chain revenue.
This challenge is open to everyone - from early-stage founders to existing startups looking to expand or pivot their products.
Your submission should clearly show how your solution integrates or reimagines X402 Payments in the context of real-world financial interaction
Tracks
Choose one - or mix and match. Each track explores a core use case of programmable payments in the age of agentic systems:
π€ Agent-to-Agent Payments
Challenge: Build a system where two or more agents autonomously trigger and settle paymentsβe.g., for usage-based services, access control, or dynamic pricing. Your product should demonstrate a working agentic payment loop with minimal human input.
π§βπΌConsumer AI Payments
Challenge: Develop an AI assistant that can make payments on a userβs behalfβwith built-in rules like spending limits, approval checkpoints, or identity verification (KYC/AML). Showcase how your product handles decision-making, compliance, and safeguards.
π’ B2B FinOps & Compliance
Challenge: Build a tool for businesses to manage real-time payments and financial operationsβlike tracking cash flow, enforcing policies, or generating audit-ready records. Highlight automation, accuracy, and visibility into payment activity.
π On-chain Commerce Primitives
Challenge: Launch a digital product or service with a built-in revenue model using X402. Think: token-gated access, real-time rev-splits, or instant payouts to contributors. Your prototype should demonstrate seamless, trustless commerce flows.
Technology partners
Gemini is Googleβs family of next-generation multimodal AI models capable of reasoning across text, images, code, video, and audio. It enables developers to build powerful AI agents and intelligent systems for real-world applications.
Google AI Studio is a browser-based environment where developers can easily prototype, test, and iterate with Gemini models before integrating them into production using the Gemini API.
Challenge:
Build intelligent agents or applications using Gemini models and Google AI Studio.
Your project should:
-
β’ Use Gemini (via Google AI Studio or the Gemini API) for reasoning, chat, or multimodal understanding
β’ Implement agent-driven or automated workflows that respond to user input, data, or context
β’ Demonstrate practical value through a working prototype or intelligent system
Recommended Gemini Models:
Participants are encouraged to select the Gemini model that best fits their agent use case:
-
Gemini Flash
Optimized for speed and low latency. Ideal for real-time applications, chat systems, and responsive agent workflows.
Gemini Pro
Designed for advanced reasoning and complex decision-making. Ideal for multi-step workflows, analysis, and enterprise-grade agents.
Gemini Flash
Optimized for speed and low latency. Ideal for real-time applications, chat systems, and responsive agent workflows.
Gemini Pro
Designed for advanced reasoning and complex decision-making. Ideal for multi-step workflows, analysis, and enterprise-grade agents.
Example Use Cases:
Gemini enables rich multimodal use cases such as:
β’ Processing and understanding documents (contracts, reports, research)
β’ Analyzing images or visual inputs for decision-making
β’ Building AI-powered internal tools (knowledge bases, assistants, automation)
Access to Technology:
Participants can explore Gemini through two main options - both include free access:
-
π Google AI Studio: A free, browser-based IDE for experimenting with Gemini. You can write prompts, test outputs, tweak parameters, and prototype without needing to set up billing or any API keys. Ideal for prototyping during the hackathon.
-
π» Gemini API: For developers ready to integrate Gemini directly into their apps. The free tier includes a monthly allowance of tokens for text and multimodal inputs/outputs, letting you test and build with Geminiβs core models at no cost. Paid usage only applies for higher-volume or production-level apps.
π Google Cloud Credits:
New Google Cloud accounts receive $300 in free Google Cloud credits, valid for 90 days.
These credits can be used across eligible Google Cloud services, including Gemini API usage, subject to Google Cloudβs standard terms and regional availability.
π Learn more about Google Cloudβs free tier
-
Note: After the free credits are used or expire, continued usage of billable services requires enabling billing on your Google Cloud account. Charges only apply if you choose to continue using paid services.
Note: After the free credits are used or expire, continued usage of billable services requires enabling billing on your Google Cloud account. Charges only apply if you choose to continue using paid services.
π How to Get Started
1. Visit the Gemini docs
2. Follow the Quickstart Guide to explore Gemini in AI Studio
4. Build and test your agent or application
π Resources:
Veea builds secure edge AI and infrastructure systems for enterprise deployments. For this hackathon, Veea introduces Lobster Trap, a deep prompt inspection and policy enforcement layer designed to help developers build secure, observable, and production-ready AI agents.
Challenge:
Build the trust layer enterprise security teams will actually sign off on.
Agents now read files, hit APIs, send messages, and trigger actions inside production systems, but the guardrails have not kept up. A single manipulated prompt can leak credentials, exfiltrate data, or trigger the wrong action while most teams still lack visibility into what their agents are doing.
Build guardrails, observability, access control, audit, or red-team tooling where policy enforcement and conversation-layer security are first-class citizens, not an afterthought.
Participants are encouraged to build projects involving:
- β’ Prompt injection and exfiltration detection
- β’ Guardrails and policy enforcement for AI agents
- β’ Access control and permission systems
- β’ Agent observability and audit tooling
- β’ Red-team and adversarial testing
- β’ Governance dashboards and explainability tooling
- β’ Multi-agent security frameworks
Bonus points for measurable risk reduction, blocked attacks, caught exfiltration, declared-versus-detected intent mismatches, or audit trails a regulator could read.
Technology
Lobster Trap is a deep prompt inspection (DPI) proxy that sits between your agents and any OpenAI-compatible LLM backend, enforcing P4-style firewall rules on every prompt and response without changing your application code.
Key capabilities:
- β’ Drop-in deployment: Works with OpenAI-compatible APIs including Ollama, vLLM, llama.cpp, OpenAI, Anthropic, and Gemini via proxy.
- β’ Real DPI, not vibes: Extracts structured metadata including intent category, risk scoring, credentials, PII, injections, exfiltration patterns, target paths, domains, and risky commands.
-
β’ Programmable policy: YAML-based rules with first-match-wins logic and actions including
ALLOW,DENY,LOG,HUMAN_REVIEW,QUARANTINE, andRATE_LIMIT. -
β’ Bidirectional metadata via
_lobstertrap: Supports declared-versus-detected intent inspection and audit reporting. - β’ Production-grade tooling: Includes real-time dashboards, audit logs, CLI utilities, and lightweight static Go binary deployment.
How Veea would like it presented: Lobster Trap is the floor, not the ceiling. Use it as the trust layer your project builds on top of, including policy packs for HIPAA, SOC2, or finance, drift monitoring, multi-agent permission systems, governance dashboards, and enterprise security workflows.
Access to Technology:
Lobster Trap is MIT-licensed and free.
Three ways to start:
-
β’ Standalone: Clone, run
make build, then start with./lobstertrap serve. - β’ Inside Native.Builder: Already packaged into lablab's Native.Builder.
- β’ Pre-built static binaries: Linux, Windows, and macOS builds with no Go toolchain required.
No API keys, signups, rate limits, or cloud dependency required.
π Resources:
β’ Repo (MIT)
β’ Quickstart & full policy reference (README)
β’ Default policy to fork: configs/default_policy.yamlβ’ Built-in adversarial test suite: ./lobstertrap testβ’ Single-prompt debugger: ./lobstertrap inspect "<your prompt>"β’ TerraFabric
Mentor support: Veea engineers active in the lablab Discord throughout the build phase for policy review, integration help, and architecture questions.
Prizes
Special partner awards for standout AI builders.
Best use of Gemini
Awarded to the top projects building with Gemini.
1st Place
$5,0002nd Place
$3,0003rd Place
$2,000Agent Security & AI Governance Track Winner
Enterprise deployment, visibility, and ecosystem opportunity prize.
Veea DevKit on NVIDIA DGX Spark
Veeaβs edge AI computing platform running on your own personal AI supercomputer, plus TerraFabric pilot access.
Technical Writeup Feature
Co-authored publication amplified across Veeaβs official channels.
Direct Intro to Engineering Team
Collaboration, pilot, networking, or hiring opportunities with Veea.
Stage Recognition
Recognition at AI & Big Data Expo North America in front of 8,000+ attendees.
What to submit?
π Basic Information
β’ Project Title
β’ Short Description
β’ Long Description
β’ Technology & Category Tags
πΈ Cover Image and Presentation
β’ Cover Image
β’ Video Presentation
β’ Slide Presentation
π» App Hosting & Code Repository
β’ Public GitHub Repository
β’ Demo Application Platform
β’ Application URL
For further details and guidance, please visit Submission Guidelines
Hackathon Details
Join lablab.ai hackathon and innovate using the latest models in the market. Discover all the relevant details below.
ποΈ Where and when
The start date of the hackathon is mentioned according to the date specified on the hackathon page, cover and schedule. The hackathon will take place on the lablab.ai platform and lablab.ai Discord server.
π¦ΈπΌββοΈ Who can join?
Everyone is welcome to participate, regardless of previous AI or coding experience! We encourage anyone with a passion for AI or an interest in exploring how it can be used in their field to join.
π What about teams?
If you don't have a team, don't worry! You can connect with other participants from all over the world on our dashboard or Discord server. We also recommend checking out our Discord server to find teammates and bounce around ideas. You can join the server here.
π οΈ How to participate
The hackathon will take place online on lablab.ai platform and lablab.ai Discord Server. Please register for both in order to participate. To participate click the "Enroll" button at the bottom of the page and read our Hackathon Guidelines and Getting Started Guide.
π§ Get prepared / Use Lablab.ai to Learn About AI
To get ready for the hackathon, visit our AI Tech pages and read up on all the available technologies. You can also check out our tutorials page for more information on how to use them. Get a head start on your project by using the resources on lablab.ai
Speakers, Mentors & Judges
Event Schedule
- To be announced
Submitted concepts, prototypes and pitches
Submissions from the teams participating in the Transforming Enterprise Through AI event and making it to the end π
.png&w=3840&q=75)
.png&w=3840&q=75)
.png&w=3840&q=75)
.png&w=3840&q=75)
.png&w=3840&q=75)
.png&w=3840&q=75)
.png&w=3840&q=75)
.png&w=3840&q=75)
.png&w=3840&q=75)
.png&w=3840&q=75)
.png&w=3840&q=75)
Teams: Transforming Enterprise Through AI
Check out the roster and find teams to join
.png&w=3840&q=75)
Next Hackathon
Band of Agents Hackathon
Starts Jun 12, 2026