BuidlAI Hack 2025

DEFAULT

We live in the world of Agents. From daily assistance to professional tasks, it is no surprise to witness the use of AI to enhance the performances. BUIDLAI Hackathon team believes that the pace of growth will accelerate in the foreseeable future and would like to invite the builders who will realize a better future through the use of AI Agents. For this reason, we invite every type of AI Agent projects including infrastructure, E-commerce, Education, Finance, and other sectors that presents the state of the art technological advances and valid use cases.

General Track will be broken down into two rounds

- Seed Round: All projects submitted to BUIDLAI Hackathon will be eligible for the Seed Round. You are required to pitch the project to the judges onsite in order to be considered to the next Round
- Finalist Round: Top seven teams will be selected through Seed Round for the final pitch. Selected teams will be pitching on stage for the final judgement. Three teams among the finalists will be selected as the winners of the hackathon

General Track

Seed Round

First Prize

Second Prize

Third Prize

NEAR AI

## **Description of the Bounty**

**What is the challenge?**

Build innovative, autonomous AI agents that solve real-world problems using the NEAR AI Agent Network. Your agent should demonstrate exceptional problem-solving capabilities, benchmark performance, and practical utility within the NEAR ecosystem.

The NEAR AI Agent Network allows you to create, deploy, and manage autonomous AI agents that can:

* Interact with users through natural language  
* Access and use a variety of tools to accomplish tasks  
* Leverage cutting-edge AI models  
* Maintain user privacy and data ownership

Choose one of these implementation paths:

* **Intelligent Data Analysis Agent**: Process complex datasets, generate insights, and present findings through visualizations and explanations  
* **Knowledge Management Assistant**: Organize information, answer questions, and maintain a knowledge base that improves over time  
* **Tool Integration Specialist**: Create an agent that effectively integrates multiple external tools to solve complex problems

## **Requirements**

### **Technical Requirements**

* Agent must be hosted on the NEAR Agent Hub ([https://app.near.ai/agents](https://app.near.ai/agents))  
* Code must be hosted in a public GitHub repository  
* Project must include quantitative benchmarks for agent performance

### **Submission Requirements**

* Complete GitHub repository with:  
  * Project name and one-line description  
  * Problem statement and solution approach  
  * Source code with clear organization  
  * Detailed README.md with setup instructions  
  * Clear evaluation metrics   
  * Technical architecture overview  
  * Future development roadmap  
* 3-5 minute demo video showing the implementation  
* Link to deployed agent on NEAR Agent Hub

## **Judging Criteria**
1. **Technical Innovation** 
   * Novel problem-solving approaches
   * Originality of use case
   *  Exploration of new AI-agent capabilities
   * Weigh: 25%
2. **Implementation Quality** 
   * End-to-end functionality 
   * Code quality and architecture
   * Performance and reliability
   * Benchmark results
   * Weigh: 25%
3. **User Experience** 
   * Interface design and accessibility
   * Ease of use
   * Documentation quality 
   * Weigh: 20%
4. **AI Integration Creativity** 
   * Thoughtful application of AI capabilities
   * Effective agent reasoning and decision-making
   * Quality of autonomous operations 
   * Weigh: 15%
5. **Technical Robustness** 
   * Security considerations
   * Scalability approach
   * Error handling and edge cases
   * Weigh: 15%

## **Bounty Amount and Allocation**

**Total Prize Pool: $5,000 USDT**
* 🥇 **First Place**  $2,500 USDC 
* 🥈 **Second Place** $1,500 USDC 
* 🥉 **Third Place**  $1,000 USDC 

##   **References for Hackers**

### **NEAR AI Resources**

* [NEAR AI Documentation](https://docs.near.ai/)  
* [NEAR AI Agent Hub](https://app.near.ai/agents)  
* [Developer Support Channel](https://t.me/+QrQgzK5pQaA5OTRh)  
* [Agent Quickstart Guide](https://docs.near.ai/agents/quickstart)  
* [Agent Tools](https://docs.near.ai/agents/tools)

### **AI Models Integration**

* [NEAR AI Models](https://docs.near.ai/models)  
* [Benchmarks](https://docs.near.ai/models/benchmarks)  
* [Fine-Tuning](https://docs.near.ai/models/fine-tuning)

### **Community Resources**

* [NEAR AI Community](https://docs.near.ai/community)  
* [NEAR AI Developer Hub](https://docs.near.ai/#developer-hub)

## **Important Dates**

* **Hackathon Start**: April 11, 2025 (Friday afternoon)  
* **Submission Deadline**: April 13, 2025  
* **Winners Announcement**: April 13, 2025 (Closing ceremony)

## **Contact Information**

For questions about this bounty, please reach out to the organizers at the event or join the [Developer Support Channel](https://t.me/+QrQgzK5pQaA5OTRh). 



NEAR AI Bounty for Buidl AI

Good Near Chain Research

News Agent

Near AI Agent For CEO

Psych

Creative

Nethermind

## Prize Pool

$5000 USD

## Challenge Overview

Are you curious about the intersection of Ethereum and Artificial Intelligence? This bounty invites you to build a compact but impactful project that showcases how AI-powered agents can interact meaningfully with blockchain technologies. Whether it’s a conversational tool that interprets on-chain data or an intelligent system that executes transactions, the focus is on practical learning through building.

**Possible Directions:**

- **DeFAI:** An agent which automatically performs onchain operations based on complex real world signals. We are not looking for simple DEX interactions or yield farming; we want to see a sophisticated strategy being ideated by agents and executed onchain. We are looking for originality of ideas combined with seamless UI rather than technical perfection.
- **Agentic Information Retrieval**: OpenAI just announced Operator. This allows Agents to use information which is retrieved without the direct usage of an API. In an agentic world we imagine that information can be retrieved from any source: a webstite, an API, a smart contract, or a sensor. Build a way for Agents to retrieve, parse and comprehend arbitrary information stowed away in an arbitrary data silo of which the schema and information architecture is unknown to the code implementation, so that this data can be used by another agentic system.

These are just a few starting points— There are **no specific requirements or constraints**—you’re free to build **anything** that connects AI and Ethereum. Whether it’s a simple interface that interprets on-chain data or an intelligent system that takes autonomous action, we’re looking for projects that push boundaries, explore new ideas, and bring agentic systems to life.

## Evaluation Framework

1. **Innovation (30%)** \- Does your concept explore uncharted territory in the AI\-blockchain domain?
2. **Implementation Quality (25%)** \- Can your application or prototype demonstrate complete end\-to\-end functionality?
3. **Experimentation Value (15%)** \- Evidence of exploration and documented learning processes throughout development
4. **AI Integration Creativity (15%)** \- Thoughtful application of AI capabilities\, even if modest in scope
5. **Interface Design (15%)** \- Accessibility and clarity of the project for potential users

## Award Structure

* **First Place:** $2,500
* **Second Place:** $1,500
* **Third Place:** $1000

## Submission Guidelines

* Must incorporate at least one AI component (natural language processing, autonomous decision-making, etc.) that interacts meaningfully with with an Ethereum (or EVM-compatible) smart contract.
* Include a concise demonstration video showcasing the project's functionality
* Provide a deployed URL for project testing (MANDATORY - submissions without functional deployment links will not be considered)
* Submit complete source code via GitHub repository

## Development Resources

**Note:** These are useful resources for beginners, please choose any AI stack you prefer.

### AI Agent Frameworks

* **Conceptual Foundations**:
    * [Anthropic's Agent Building Guide](https://www.anthropic.com/research/building-effective-agents)
    * [MCP](https://www.anthropic.com/news/model-context-protocol)
* **Retrieval-Augmented Generation (RAG)**:
    * [LangChain](https://www.langchain.com/)
    * [LlamaIndex](https://www.llamaindex.ai/)
    * [Haystack](https://haystack.deepset.ai/)
* **LLM Orchestration Tools**:
    * [LangGraph](https://www.langchain.com/langgraph)
    * [CrewAI](https://www.crewai.com/)
    * [AutoGen](https://github.com/microsoft/autogen)
    * [Swarm](https://github.com/openai/swarm)
* **Autonomous Agent Implementations**:
    * [Eliza (JavaScript/TypeScript)](https://github.com/elizaOS/eliza)
    * [ZerePy (Python)](https://github.com/blorm-network/ZerePy)
    * [G.A.M.E. Framework](https://whitepaper.virtuals.io/developer-documents/game-framework)
    * [Virtuals](https://app.virtuals.io/)

### Ethereum Development

* [Official Ethereum Developer Portal](https://ethereum.org/en/developers/)
* [Ethereum Smart Contract Best Practices](https://consensys.github.io/smart-contract-best-practices/)
* [Foundry Development Framework](https://book.getfoundry.sh/)
* [Ethers.js Documentation](https://docs.ethers.org/)
* [OpenZeppelin Contract Library](https://docs.openzeppelin.com/)

### AI Model Access

* [OpenAI API](https://platform.openai.com/docs/overview)
* [Anthropic Claude API](https://docs.anthropic.com/claude/reference/getting-started-with-the-api)
* [Hugging Face Models](https://huggingface.co/models)
* [Cohere API](https://docs.cohere.com/docs)
- [OpenAI API](https://platform.openai.com/docs/overview)
- [Anthropic Claude API](https://docs.anthropic.com/claude/reference/getting-started-with-the-api)
- [Hugging Face Models](https://huggingface.co/models)
- [Google Gemini](https://ai.google.dev/)
- [smolagents](https://www.notion.so/1ac360fc38d08098ac3cde1c73973909?pvs=21)
- [Open Source Deep Research](https://huggingface.co/blog/open-deep-research)

### **Building Effective AI Agents** [https://www.youtube.com/watch?v=7c3wnyrfadk&t=1s](https://www.youtube.com/watch?v=7c3wnyrfadk&t=1s)
    * **Embedding, Vector Databases & RAG Infrastructure**
        * **Chroma** ([GitHub](https://github.com/chroma-core/chroma)): Lightweight, easy-to-use vector database.
        * **Pinecone** ([Website](https://www.pinecone.io/)): Scalable cloud-native vector DB.
        * **FAISS** ([GitHub](https://github.com/facebookresearch/faiss)): Fast CPU/GPU vector similarity search.

    ### **Evaluation & Prompt Engineering**
    * **OpenAI Evals** ([GitHub](https://github.com/openai/evals)): Framework for evaluating and benchmarking AI agents.
    * **HELM Benchmark** ([GitHub](https://github.com/stanford-crfm/helm)): Holistic evaluation of language models.
* [https://youtu.be/SQzTSNurehU?feature=shared](https://youtu.be/SQzTSNurehU?feature=shared)

Create Your Agentic Future

First Place

Second Place

Third Place

Gensyn

### **Problem Statement**

The future of machine learning is both decentralised and fragmented. Our current monolithic models will be replaced by sharded parameters existing over every device in the world. Gensyn’s research is looking to discover what the different paths towards this future are, and we believe reinforcement learning (RL) post-training is one of the most effective first approaches.

### **Bounty 1: UI**

In this bounty, you are expected to create an alternative dashboard on top of RL Swarm and/or the on-chain data exposed by swarm nodes when they interact with smart contracts.

Suggestions: visualise different data than the existing swarm UI, visualise data/progress from the point of view of a single node, track performance of a model through multiple swarms.

Prizes

1. Winner: $750
2. Runner-up: $250

### **Resources**

- [Gensyn Research](https://www.gensyn.ai/research)
- [RL Swarm - Read More](https://www.gensyn.ai/articles/rl-swarm)
- [RL Swarm - Report](https://github.com/gensyn-ai/paper-rl-swarm/blob/main/latest.pdf)

POC: chloe@gensyn.ai

Bounty 1: UI

Winner

Runner Up

### **Problem Statement**

The future of machine learning is both decentralised and fragmented. Our current monolithic models will be replaced by sharded parameters existing over every device in the world. Gensyn’s research is looking to discover what the different paths towards this future are, and we believe reinforcement learning (RL) post-training is one of the most effective first approaches.

### **Bounty 2: On-chain**

In this bounty, you are expected to create a set of smart contracts to extend the functionality of RL Swarm or add additional functionality making use of the RL Swarm data.

Suggestions: create a prediction market over swarm leaderboards, add new voting systems within the swarm coordination contracts, add tradeable ownership to individual models.

Prizes

1. Winner: $1,500
2. Runner-up: $500

### **Resources**

- [Gensyn Research](https://www.gensyn.ai/research)
- [RL Swarm - Read More](https://www.gensyn.ai/articles/rl-swarm)
- [RL Swarm - Report](https://github.com/gensyn-ai/paper-rl-swarm/blob/main/latest.pdf)

POC: chloe@gensyn.ai

Bounty 2: On-chain

### **Problem Statement**

The future of machine learning is both decentralised and fragmented. Our current monolithic models will be replaced by sharded parameters existing over every device in the world. Gensyn’s research is looking to discover what the different paths towards this future are, and we believe reinforcement learning (RL) post-training is one of the most effective first approaches.

### **Bounty 3: Off-chain (swarm)**

In this bounty, you are expected to create and initialise a custom RL Swarm around a specific theme or dataset by modifying the swarm code itself and possibly the coordination smart contracts.

Suggestions: create a swarm that uses different data modalities, create a swarm to solve coding problems, create a swarm in a language other than english.

Prizes

1. Winner: $1,500
2. Runner-up: $500

### **Resources**

- [Gensyn Research](https://www.gensyn.ai/research)
- [RL Swarm - Read More](https://www.gensyn.ai/articles/rl-swarm)
- [RL Swarm - Report](https://github.com/gensyn-ai/paper-rl-swarm/blob/main/latest.pdf)

POC: chloe@gensyn.ai

Bounty 3: Off-chain (swarm)

Story

**Prize Pool**

* Top prize: $4,000  
* Second prize: $2,000  
* Third prize: $1,500

**Description**  
Let’s build a sustainable AI future, where content from humans and models is fairly attributed, licensed and monetized. Integrate your agent, model or app with Story to compete for this bounty. From IP registration to seamless licensing and royalty management, Story lays the foundation for AI collaboration and agentic knowledge transactions.

**Requirements**

1. Projects must integrate with Story, i.e. integrate with the Story SDK (or smart contracts directly). At a bare minimum, apps should at least write (and read) some form of on-chain state, related to IP.  
2. Judging will be based on how thoughtful the integration with Story is. For example, an agent that is registered as IP, licenses data from other agents, and licenses its own training data/outputs will be prioritized more than an agent that simply registers its image generation output, or one that doesn’t register at all\!  
3. Top prizes will prioritize submissions that extensively interact with IP assets on chain.

**References**

* Learn more about Story at: docs.story.foundation  
* Get started with story at: https://docs.story.foundation/docs/quickstart

Agent on Story

Saga ⛋

Description of the bounty
- Build an innovative AI agent or agentic tooling solution with onchain integration on a Saga chainlet. 
- Frameworks like Eliza, VirtualOS with onchain integration
- DeFAI agents which are trading onchain or engage with CT like [@AIXBT](https://t.me/AIXBT)

- Social platform for agents to chat and interact 
- Gaming agents which engage in web3 games

Requirements
- Smart contract deployed on a Saga Chainlet
- Projects must have an accessible UI/frontend, and a GitHub repository with a descriptive README which includes a short description of your submission and why you chose Saga chainlet

Judging criteria
- Innovation
- Use of AI
- Usage of Saga chainlet
- Degen ratio

Breakdown of bounty amount and allocation (how much is 1st/2nd/3rd prize etc)
- 10k (5/3/2)

References for hackers
- Developer Resources: [https://www.saga.xyz/academy](https://www.saga.xyz/academy)
- Developer Documentation: [https://docs.saga.xyz/introduction/saga-overview](https://docs.saga.xyz/introduction/saga-overview)

AI Agent on Saga

1st Prize

2nd Prize

3rd Prize

Eigen Layer

**Best new AI AVS**  
**1st Place \- $1,500**  
**2nd Place \- $500**   

Criteria:  
Build and open source a new and unique AVS which helps to power other AI related applications. This can be a part of your stack or fall into one of these categories.   
Developers can build from scratch, fork "Hello World", or build with products like Othentic ([https://docs.othentic.xyz/main](https://docs.othentic.xyz/main)) , Layer ([https://www.layer.xyz/\#layer-stack](https://www.layer.xyz/#layer-stack)) , AltLayer, or Tangle / Gadget

Success:  
A successful AVS implementation should demonstrate:

1. A unique and valuable service that leverages EigenLayer's security \- for example, a new type of a novel verifiable AI solution, or a specialized computation service  
2. Technical implementation requirements:  
   * Implementation of the core AVS interfaces and middleware  
   * Clear documentation of the service architecture and deployment process  
   * Working proof-of-concept that demonstrates the service in action  
3. Desired features:  
   * Complete integration with EigenLayer's testnet  
   * Proper handling of operator registration and management

Concrete examples of successful AVS could include:

* A decentralized ML model training and serving platform

Extra credit will be given to projects that demonstrate novel use cases, high potential for real-world adoption, and thorough testing and documentation.

Judging

1. Technical Implementation (35%)  
   * Code quality and best practices  
   * Completeness of implementation  
   * Successful integration with EigenLayer testnet  
2. Innovation & Utility (35%)  
   * Uniqueness of the AVS use case  
   * Potential impact and market need  
   * Technical feasibility  
3. Documentation & Presentation (30%)  
   * Quality of technical documentation  
   * Clear explanation of the service architecture  
   * Deployment and testing instructions

Resources

* EigenLayer Resources [https://eigen-labs.notion.site/Eigen-Developer-Resources-11c13c11c3e0805fbd48e85799d14e12?pvs=4](https://eigen-labs.notion.site/Eigen-Developer-Resources-11c13c11c3e0805fbd48e85799d14e12?pvs=4)  
* EigenLayer Docs [https://docs.eigenlayer.xyz/](https://docs.eigenlayer.xyz/)  
* EigenLayer Hello World [https://github.com/Layr-Labs/hello-world-avs](https://github.com/Layr-Labs/hello-world-avs)

Use cases

Anything novel  
Verifiable AI  
AI Prediction markets  
Verifiable agential infrastructure  
Privacy-Preserving Data Oracle   
Cross-Chain Bridge Validation   
Decentralized ML Model Inference 

Best new AI AVS

1st Place

2nd Place

**Best new AI Application or AI Agent built with an AVS**
**1st Place \- $1,500**   
**2nd Place \- $500**

Description

Build an AI application or agent that leverages EigenLayer or an EigenLayer AVS, such as implementing verifiable inference, verifiable agent, AI model governance system, AI agent reputation system, or anything else you can imagine.

What is required to complete this bounty

* Include a working implementation of an AI application or agent using an EigenLayer AVS, or a new AI-focused AVS   
* Be open source with complete documentation  
* Include a functional prototype that can be cloned and run in some way, or used via web or API interface  
* Provide clear explanation of how the app or agent works

What does success look like?

Novelty

* A unique and interesting use case, agent, or application that provides real-world entertainment or value to the user

Simple developer onboarding

* Well-documented readme that is easy to understand  
* Clear demonstration of AI capabilities  
* Codebase should be able to be cloned and run to test out 

Innovation & Impact

* Practical or entertaining use case  
* Potential for real-world adoption  
* Clear / easy to understand value proposition or use case

Nice to haves:

* Command-line interface or web-based dApp with intuitive user interactions or instructions  
* Public demo URL for browser-based applications or containerized package for CLI tools  
* Clear visualization of the verification process and AI system outputs

How are you judging?

* Technical Implementation (40%) \-  completeness, a working prototype  
* Innovation & AI Impact (30%) \- Uniqueness, practicality, and real-world potential  
* Documentation & Testing (20%) \- Technical docs and deployment guides  
* User Experience (10%) \- Interface quality and ease of use

What is the impact on your organization by a team working on this bounty?

* Successful AI projects will help establish EigenLayer as a key infrastructure for verifiable and trustworthy AI applications in Web3, advancing the frontier of what's possible with on-chain / verifiable AI.

Resources and links

* Hyperbolic Labs [https://hyperbolic.xyz/](https://hyperbolic.xyz/)  
* Gaia [https://www.gaia-ai.eco/](https://www.gaia-ai.eco/)  
* EigenLayer Resources [https://eigen-labs.notion.site/Eigen-Developer-Resources-11c13c11c3e0805fbd48e85799d14e12?pvs=4](https://eigen-labs.notion.site/Eigen-Developer-Resources-11c13c11c3e0805fbd48e85799d14e12?pvs=4)  
* EigenLayer Docs [https://docs.eigenlayer.xyz/](https://docs.eigenlayer.xyz/)  
* EigenLayer Hello World [https://github.com/Layr-Labs/hello-world-avs](https://github.com/Layr-Labs/hello-world-avs)  
* Opacity Eliza Plugin [https://x.com/dabit3/status/1877056923092267066](https://x.com/dabit3/status/1877056923092267066)

* What are examples of use cases you’re trying to solve?  
* Verifiable AI Model Inference \- Validate and verify AI model outputs on-chain, ensuring transparency and correctness of AI predictions  
* AI Agent Reputation System \- Track and verify AI agent actions and build reputation scores based on their performance and reliability  
* AI Model Governance \- Create a decentralized system for validating AI model updates and ensuring compliance with safety parameters  
* Secure Model Access Control \- Manage and verify access to AI models while maintaining an immutable audit trail of all interactions  
* AI Content Attribution \- Validate the authenticity and provenance of AI-generated content to combat misinformation  
* Cross-Chain AI Verification \- Enable consistent and verified AI computations across different blockchain networks  
* AI Safety Monitoring \- Create verifiable checks and balances for AI system behaviors and safety compliance  
* Decentralized Model Training Verification \- Validate the training process and data quality for AI models in a decentralized way

Best new AI Application or AI Agent built with an AVS

PIN AI

# **🚀 Hackathon Challenge: Personalized AI Agents for the PIN AI Ecosystem**

## **🎯 Description of the Bounty**

AI agents are no longer just research experiments—they’re powerful, real-world tools automating workflows, enhancing decision-making, and integrating seamlessly into daily life. PIN AI is building a future where your AI agent effortlessly arranges your travel plans, manages your finances, or curates personalized shopping experiences based on your unique preferences and habits.

The goal of this hackathon is to **build innovative AI agent services capable of leveraging personal data.** Your AI agent service will be deployed on the PIN protocol and will be able to receive service requests from real users personal AIs (for example “plan my next trip”, “help me buy the best gift for my partner's birthday”, “find my next job opportunity to advance my career”, etc). A sandbox development environment is available to help you test your AI agent services against a synthetic personal AI with user data and personality, along with demos and examples to inspire your creations (see resources below).

Join builders, researchers, and innovators pushing the boundaries of AI automation, memory, reasoning, and personalization. Build cutting-edge AI Agents within the PIN AI ecosystem\!

## **✅ Requirements**

* All attendees must contribute code to a public GitHub repository.  
* Teams must consist of up to four members and can be formed during the hackathon.  
* Teams must demonstrate a Personal AI \<\> AI Agent interaction similar to [this demo](https://drive.google.com/file/d/1ITCwFLerjHspFJ7bohU2jU3h8QdGTNVE/view?usp=sharing).

## **🏅 Judging Criteria**

1. **Functionality (50%)**: AI Agents should fully serve Personal AI in end-to-end flows—exchanging information, negotiating, and processing real transactions. For example, a food-ordering agent should place actual delivery orders, and a flight-booking agent should successfully book flights and confirm bookings.  
2. **Registration (25%)**: All AI Agents must be registered in the PIN AI Agent Registry.  
3. **Innovation (25%)**: Does your agent deliver a service that leverages personal data in a way that is novel and provide a concrete service that is not possible without that personal data or from traditional AI tools?

## **💰 Prizes**

* 🥇 **1st Place**: 1500 USDT \+ PIN AI community rewards  
* 🥈 **2nd Place**: 1000 USDT \+ PIN AI community rewards  
* 🥉 **3rd Place**: 500 USDT \+ PIN AI community rewards

## **💡 Building Inspiration**

🔹 **Life OS Assistant** A persistent AI companion that learns from interactions, manages schedules, organizes tasks, and adapts to evolving routines.

🔹 **AI Recruiter & Job Matching Assistant** A smart AI agent that effortlessly connects job seekers to their dream roles by matching skills, interests, and career aspirations with current job openings.

🔹 **AI Dating Matchmaker 💖** A personal dating assistant leveraging personality profiles, interests, and interactions to find and arrange meaningful matches and memorable dating experiences.

🔹 **AI Style & Fashion Curator 👟** An AI that learns personal style preferences, tracks fashion trends, and curates personalized outfit recommendations and seamless shopping experiences.

🔹 **AI Therapist & Wellness Coach** An AI dedicated to mental well-being—offering mood tracking, guided meditation, and habit-building suggestions grounded in cognitive science.

🔹 **Personalized AI Financial Advisor** An AI-powered wealth manager that tracks spending, optimizes investments, and provides real-time financial insights.

## **📚 Helpful Resources**

1. **Background Information on PIN AI**

Official Website :  
[https://www.pinai.io/](https://www.pinai.io/)

Video demo :   
[https://drive.google.com/file/d/1MWTqqDSAoZHIxk5JiWoKjPdKnThAWBIu/view?usp=sharing](https://drive.google.com/file/d/1MWTqqDSAoZHIxk5JiWoKjPdKnThAWBIu/view?usp=sharing)  
[https://drive.google.com/file/d/1k6JvB6gvOvrP6thGtVm2YWHHukJ6l5Gc/view?usp=sharing](https://drive.google.com/file/d/1k6JvB6gvOvrP6thGtVm2YWHHukJ6l5Gc/view?usp=sharing)  
[https://drive.google.com/file/d/1ITCwFLerjHspFJ7bohU2jU3h8QdGTNVE/view?usp=sharing](https://drive.google.com/file/d/1ITCwFLerjHspFJ7bohU2jU3h8QdGTNVE/view?usp=sharing)

Google form link for participants info : [https://docs.google.com/forms/d/e/1FAIpQLSdM41iInCdA7qc8fhMSVtsu1wi1HwQSUASo7e8siivdjpYDgQ/viewform](https://docs.google.com/forms/d/e/1FAIpQLSdM41iInCdA7qc8fhMSVtsu1wi1HwQSUASo7e8siivdjpYDgQ/viewform)

2. **Resources for Agent Development**

Agent SDK & Examples (Please follow the tutorial here): 

* [https://github.com/PIN-AI/pinai\_agent\_sdk](https://github.com/PIN-AI/pinai_agent_sdk)  
* [https://pypi.org/project/pinai-agent-sdk/](https://pypi.org/project/pinai-agent-sdk/)

PIN Agent platform :   
[https://agent.pinai.tech/](https://agent.pinai.tech/)

Agent marketplace demo (Byte AI) :  
[https://github.com/PIN-AI/pinai\_agent\_marketplace\_demo](https://github.com/PIN-AI/pinai_agent_marketplace_demo).    
[https://x.com/Byte\_\_AI](https://x.com/Byte__AI)

dev-guide  
[https://drive.google.com/file/d/1mO0UuGliloJDW6ilLh8ragsQWO4gqHl6/view?usp=drive\_link](https://drive.google.com/file/d/1mO0UuGliloJDW6ilLh8ragsQWO4gqHl6/view?usp=drive_link)

3. **Resources for Agent Registration**

Get some test $ETH for free using Base Sepolia faucet :   
[https://www.basefaucet.xyz/](https://www.basefaucet.xyz/)

Agent registry contract :   
[https://sepolia.basescan.org/address/0xD2004b20B39A6b4397df87dadDaEFB0aEfe32089](https://sepolia.basescan.org/address/0xD2004b20B39A6b4397df87dadDaEFB0aEfe32089)



Personalized AI Agents for the PIN AI Ecosystem

🥇 1st Place

🥈 2nd Place

🥉 3rd Place

RootStock

**BUIDL Asia x Rootstock | Bounty Details**

# **Description of the bounty (What is the challenge?)**

[*🏆*](https://emojipedia.org/trophy) ***Best AI Dapp on Rootstock***   
*Build high quality, impactful and innovative AI Dapps on Rootstock. Ideas can include:* 

* **DeFi Automation Agents:** AI that optimizes yield farming, trading, or risk management in DeFi protocols.  
* **Smart Contract Assistants:** Agents that help write, deploy, or audit smart contracts on Rootstock.  
* **Developer Productivity Bots:** Tools that guide onboarding, troubleshoot issues, or suggest improvements to dev workflows.  
* **Cross-Chain Interoperability Agents:** AI that streamlines communication or monitoring between Rootstock and other chains.  
* **Decentralized Governance Helpers:** Agents that simplify voting, predict outcomes, or support DAO decision-making.  
* **Blockchain Data Analytics:** AI tools that analyze on-chain data to provide trends, insights, or predictive analytics.  
* **Security & Fraud Detection:** AI-driven solutions to monitor transactions and detect anomalies or vulnerabilities.

# **Requirements**

1. Smart contract deployed in Rootstock and at least 2 on-chain (testnet) successful transactions.  
2. Projects must have a description and a GitHub repository with a descriptive README.

   

*The README should include:*

* Clear short one-sentence description of your submission.  
* Short description of what you integrated Rootstock with and how.  
* Short description of the team and their backgrounds.  
* Clear instructions for testing the integration.  
* Feedback describing your experience with building on Rootstock.  
* A short video demo or slide deck.

# **Judging criteria** (Scored from 1-5)

* **Presentation:** Clear communication of the problem, solution, and value proposition.   
* **Potential Impact:** The potential for commercial success, scalability, and addressing a significant market need.   
* **Application of Technology:** The demonstration of the technology’s functionality and how well it’s applied.   
* **Originality:** The uniqueness and innovation of the idea. 

# **Breakdown of bounty amount and allocation (how much is 1st/2nd/3rd prize etc)**

* 1st Prize $5000  
* Runner-up $2000 

# **References for hackers**

* [Developer Guide and Cheat Sheet](https://apple-whippet-f90.notion.site/Developer-Guide-Cheat-Sheet-196b6a39f4d8808094cfdec388a2a903)   
* We are also working on an AI guidebook that we will share closer to time 



Best AI Dapp on Rootstock

Runner-up

Upstage

- **Description of the bounty (What is the challenge?)**
    - How can we develop an AGI-powered application that efficiently processes and utilizes document-based data?
- **Requirements**
    - **Technical Requirements**
        - Using the Upstage API - DP(Document PArse) or IE(Information Extract) are mandatory
    - **Submission Requirements**
        - Product presentation deck (PDF)
            - Content to include
                - Project overview and objectives
                - Main features and technical characteristics
                - How Upstage API was utilized
                - Project innovation and social impact
                - Future development plans
        - Product demonstration video ( ~ 5min)
        - Project Github Repository
            - Submit a link to the GitHub repository containing your project's entire source code.
            - Clearly describe the project overview, installation method, and usage instructions in the [README.md](http://readme.md/) file.
                - During project evaluation, we can run and test the code.
            - Clearly indicate the sections where Upstage API is used.
- **Judging criteria**
    - **Project Impact (25)**
        - Professionalism (5)
        - Practicality/ Usability (10)
        - Potential for growth (10)
    - **Technical Implementation (25)**
        - AI Service Implementation Completion (15)
        - Application of Latest Technologies/Trends (10)
    - **Effective Use of Upstage API (25)**
        - Apply the appropriate API to achieve the objective (10)
        - Superiority of the API application results (15)
    - **Innovation and Creativity (25)**
        - Business Innovation (15)
        - Work Creativity (10)
- **Resources**
    - Upstage Edu Full Package v1. [https://upstage-ai-education.gitbook.io/upstage-edustage/basics/editor-2](https://upstage-ai-education.gitbook.io/upstage-edustage/basics/editor-2)
        - The Upstage Edu Full Package is a comprehensive educational program designed for beginners in AI learning.
        - With the **Edu Full Package**, anyone can quickly grasp the essential knowledge and usage of Upstage products with ease.
    - Upstage Edu Full Package v2. - TBU(4/6)
- **Bounty Amount**
    - First Prize: $1,500, Second Prize: $1,000, Third Prize: $500

BuidlAI Hack 2025

BuidlAI Hack 2025