Powering AI minds: OriginTrail hackathon
Do you have what it takes to revolutionize the future of AI agents?
Harness the power of the OriginTrail Decentralized Knowledge Graph (DKG) to create AI agents that push the boundaries of what’s possible. Combine the best of neural and symbolic AI to build agents with collective memory capabilities, meet like-minded pioneers, and compete for exciting prizes!
First DKG-powered agents like ChatDKG (powered by the ElizaOS DKG Plugin) and Satoshi Rakic have already demonstrated how the DKG unlocks next-level intelligence for AI agents. Now, it’s your turn to lead the charge and showcase your brilliance!
💡 How to join the hackathon?
Explore challenges: Check out the themes and prizes below.
Register to compete: Complete the Google form by Thursday, January 23, at noon.
Build your solution: Create an AI agent using the OriginTrail DKG as its brain.
Submit your work: Send your GitHub repo (with a clear README) to office@origin-trail.com by Monday, January 27, at noon.
Showcase your work: Join us for Demo day at Nova Iskra Dorćol, Belgrade, on January 27, 18:00!
The hackathon kicks off remotely, giving you the flexibility to collaborate and innovate from anywhere, with full support via Discord. Then, come to Belgrade for an unforgettable hackathon finale:
✨ What’s in store on Demo day?
Inspiring talks from the team and guest speakers.
Creative presentations of AI agent projects.
Live jury deliberations to crown the winners.
Award celebrations with the OriginTrail community.
Networking opportunities to connect with innovators and thought leaders.
Not competing? No problem! Register via Eventbrite to join the event, enjoy the excitement, and connect with the team and fellow enthusiasts.
Your next breakthrough starts here. Let’s build the future of AI agents together!
This is an in-person hackathon, only teams that will physically join the hackathon Demo day will be eligible for rewards.
🤖 Builder resources
Here's some code and learning resources to get you started:
Get started with the DKG here
Try out ChatDKG on Twitter and see how it creates memories
Join Discord and learn together with the community
Learn more about OriginTrail on our YouTube Channel
Useful code repos and guides:
Build a DKG agent with the Python SDK: https://docs.origintrail.io/dkg-v8-current-version/ai-agents/custom-dkg-python-agent
Learn how to setup & use a DKG Edge node here
🤺 Hackathon project challenge ideas
1. Social and influencer agents
Description: Develop agents to analyze, monitor, and enhance social media influence or content. These agents could provide insights on audience engagement, detect trends, or suggest strategies for content optimization. Example use of DKG: Leverage the OriginTrail DKG's advanced neuro-symbolic capabilities for robust knowledge retrieval, enabling cross-platform connectivity and context-aware insights. Connect and analyze influencer data as knowledge asset graphs (e.g., engagement metrics and content performance) for enhanced interpretability and prediction-making..
2. Agent infrastructure
Description: Create a launchpad or foundational frameworks to enable the development and deployment of agents that use the DKG (via gateway or DKG Edge Nodes) as a decentralized memory system. These tools could simplify agent creation and ensure interoperability across different platforms. Example use of DKG: Utilize the DKG as a shared neuro-symbolic repository where agents store their memories, data, and operational states. Its advanced connectivity and context-aware data modeling ensure resilience, scalability, and collaborative ecosystems. This approach allows agents to retrieve and link complex knowledge assets seamlessly, fostering efficient decentralized workflows.
3. DKG AI Agents
Description: Build agents that offer seamless interactions with the DKG, making it simple to create, query, and manage decentralized knowledge assets. These agents would abstract the complexities of DKG operations, allowing users to deploy and manage their Core and Edge nodes, create knowledge assets, and retrieve data effortlessly. Example use of DKG: Enable agents to leverage the neuro-symbolic strengths of the DKG to automate complex tasks like node management, asset creation, and data querying. With its advanced knowledge representation and inferencing capabilities, the DKG empowers agents to retrieve and synthesize meaningful insights while maintaining decentralized data integrity and transparency.
4. Trading agents
Description: Develop agents for autonomous trading in financial markets, leveraging the DKG for data integrity and transparency. These agents can analyze market trends, execute trades, and manage portfolios. Example use of DKG: Use the DKG's neuro-symbolic features to store and interlink historical trading data, market signals, and strategy insights. By enabling advanced connectivity and context-aware retrieval, the DKG facilitates tamper-proof, trusted data for informed decision-making. Its decentralized structure ensures a collaborative and transparent environment for autonomous trading agents, while enabling privacy via Edge nodes.
5. Agent swarms
Description: Develop coordinated groups of decentralized agents (swarms) that can collaborate to achieve complex goals, such as large-scale data aggregation, social media trend analysis, or interacting in some interesting ways between them, such as spawning new agents themselves etc. These agents could work together autonomously, leveraging distributed intelligence and dynamic task allocation.
Example use of DKG: Use the DKG to enable seamless communication and synchronization between agents in the swarm, leveraging its advanced neuro-symbolic capabilities for context-aware decision-making and knowledge sharing. For example, in Web3, agent swarms can analyze token distribution patterns, monitor governance proposal activity, and assess the health of decentralized autonomous organizations (DAOs) by interlinking blockchain activity with community metrics. On social media, swarms can track trending hashtags, analyze cross-platform content virality, and identify influencer collaborations. The DKG ensures these operations remain decentralized, transparent, and resistant to manipulation, providing a trustworthy framework for swarm-based intelligence.
6. Meme agents
Description: Build agents dedicated to promoting OriginTrail and the DKG through viral content such as memes. These agents could generate and share memes on platforms like Twitter and incentivize users based on engagement metrics. Example use of DKG: Utilize the DKG to store and interconnect the most popular memes with cultural and temporal trends, creating a decentralized, contextually enriched archive of viral content. The DKG's neuro-symbolic strengths facilitate the analysis of meme engagement patterns, enabling incentivization mechanisms rooted in verified, decentralized interaction metrics, promoting community-driven content creation and trust.
7. Wildcard
Build whatever you think is cool with Agents and the OriginTrail DKG to impress the jury. Anything goes - from graph reasoning, and agent swarms, token launching, to custom Edge Node integrations for privately run, local agents. Review the judging criteria to guide you.
🏆 Prizes
The prizes are designed to help teams bring their agents to life and move them from MVPs to actual products on the DKG. The rewards are structured as follows:
First Place: $5,000 in total value (50% in TRAC, 50% in USDC)
Second Place: $2,500 in total value (50% in TRAC, 50% in USDC)
Third Place: $1,000 in total value (50% in TRAC, 50% in USDC)
TRAC rewards are intended to cover the operational cost of your agent on the DKG and will be distributed to your agent's DKG funding contract.
To ensure the funds are used to launch and support the agents, the winners’ wallet addresses will be publicly shared. This approach encourages transparency and underscores the commitment to turning ideas into actionable, impactful solutions.
🧑⚖️ Judging criteria for the hackathon challenge
1. Effective use of the DKG (30% weight)
How well does the submission leverage the OriginTrail Decentralized Knowledge Graph?
Relevance: Does the project demonstrate correct and meaningful use of the DKG?
Alignment: Does the project generate a reasonably high number of knowledge assets on the DKG?
Depth: Are agents contributing valuable, high-quality knowledge to the DKG?
2. Innovation and creativity (20% weight)
How novel or unique is the solution?
Novelty: Does the submission introduce fresh ideas, methodologies, or applications?
Creativity: Is the approach inventive or unexpected, particularly in agent behavior, workflows, or gamification?
3. Integration and ecosystem impact (15% weight)
How well does the project integrate with other tools, platforms, frameworks, or blockchain ecosystems?
Interoperability: Does it showcase meaningful integration with other tools, frameworks, APIs, blockchain products, etc.?
Ecosystem Contribution: Could this project strengthen the broader OriginTrail ecosystem?
4. Usability and viability (35% weight)
Does the submission present a compelling, practical agent, agentic swarm or agentic product that has the potential to grow into its own product or living agent on the DKG? Is there a plan on how this agent can continue being useful post-hackathon?
Real-World Potential: Is your agent or agentic product MVP live on DKG mainnet? Does it show traction, can it grow into a valuable product?
Functionality: Does it work effectively with minimal bugs or errors?
User Experience: How easy, pluggable, and versatile is your agentic product for use? Can it tap into different agents’ memories, speak via Telegram, Discord, Twitter, and video at the same time? The more versatile, the better
💡 Engage with the community
Feel free to get in touch with the OriginTrail community and core developers for any assistance through Discord. We’ve also opened a dedicated channel exclusively for hackathon participants where you can meet each other, look for team members if you need any, share ideas, and meet the OriginTrail team. (applicants will automatically be added to the channel)
Have fun hacking and looking forward to seeing you in Belgrade!
Last updated
