![[object Object] icon](/_next/static/media/telegram.004abf19.svg){kind=small}
![[object Object] icon](/_next/static/media/globe.47e802f2.svg){kind=small}
Loading Youtube Video
Watch And Earn
+
Introduction to Pond
- Pond is a crypto-native AI model layer purpose-built to develop, deploy, and monetize machine learning models tailored for blockchain ecosystems. Designed to serve as the foundational AI infrastructure in Web3, Pond combines Large Language Models (LLMs) and Graph Neural Networks (GNNs) to power intelligent agents, provide structured onchain insights, and offer autonomous reasoning capabilities—all within a permissionless, composable, and community-owned framework.
Dylan’s Web3 Journey & How Pond Started
- Dylan’s entry into Web3 came through a deep-rooted interest in financial systems and machine learning. With a background spanning quantitative finance and chemical engineering, he witnessed firsthand the inefficiencies in both traditional AI deployment and onchain data access. His prior work with Spice Finance and DeFi protocols gave him a front-row seat to Web3’s fragmented data landscape. This sparked the genesis of Pond: a platform designed to enable open-source, blockchain-aligned machine learning infrastructure.
Shaping the Vision of Pond
- The vision behind Pond is to build the foundational AI layer for crypto—not a generalized LLM playground, but purpose-built models for onchain applications. Rather than simply porting traditional AI models into Web3, Pond rethinks how models are trained, owned, and executed across decentralized networks. The team believes that crypto needs agents that can reason about protocols, parse smart contracts, monitor liquidity, and even make governance decisions. Pond is where those agents are born.
How Pond Stands Out from Other AI Platforms
- Native GNN + LLM fusion for reasoning over smart contracts, graphs, and protocol data
- Model provenance & shared ownership, enabling contributors to earn through collective model training
- Deployed for composability, allowing models to become onchain services, callable by dApps, DAOs, or agents
- Open and permissionless, meaning anyone can plug into Pond to train or use models
- This makes Pond a first mover in crypto-specific AI infrastructure—providing not just tools, but a network effect for models and their contributors.
Funding Strategy for Pond
- Pond raised $7.5M in its seed round (2024) led by Archetype, alongside investors like Illia Polosukhin and Suji Yan. But more than capital, the round was about aligning with long-term believers in crypto’s AI moment. The team avoided hype-driven fundraising and instead focused on credible operators who understood model infrastructure, blockchain scaling, and the upcoming wave of AI agents.
Key Metrics During Pond’s First Raise
- A robust model pipeline (LLMs, GNNs, and hybrid architectures)
- Growth of its developer and contributor community
- Deep integrations with onchain protocols and tooling
- Demand for AI-powered agents from top Web3 teams and infra providers
Pond’s Developer Ecosystem & How to Get Involved
- Exploring the Pond SDK and model repo
- Joining hackathons, grants, and model bounties
- Contributing to data ingestion, GNN-based tooling, or agent wrappers
- Collaborating with the Pond core team on training or deployment
Advice for Web3 AI Founders (From Dylan)
- Start with crypto-native needs. Don’t just copy-paste traditional AI tools—understand the data, actors, and trust model of blockchains.
- Prioritize composability. Make your AI modular and callable by other onchain apps or agents.
- Own your models. Web3 is about shared upside—design systems where contributors get to benefit.
- In Web3, it’s not about scaling fast—it’s about scaling credibly with the right primitives.
Looking Ahead: What Pond Believes Will Fly in 2025
- Autonomous onchain agents with native reasoning and protocol awareness
- DePIN-AI hybrids, where data from decentralized networks feeds AI systems
- Collaborative AI ownership, where model contributors earn ongoing rewards
- Onchain AI marketplaces, with models and agents offered as public goods or monetized services
Your Progress
Current Progress:
0/0
Pond is a decentralized AI model layer designed specifically for the crypto ecosystem. It allows developers, data scientists, and researchers to collaboratively build, own, and monetize AI models using on-chain blockchain data. It’s structured to function like an open-source AI development network, optimized for Web3.
0/0
Step 1/3
Watch Podcast
Watch pond ai podcast on inflection and earn rewards
Earn
+
0 Participants
Step 2/3
Complete Pond Quiz
Test your knowledge about pond's decentralized ai platform and its capabilities.
Earn
+
0 Participants
Step 3/3
Follow Pond on Twitter
Follow pond on twitter and earn rewards.
Earn
+
0 Participants
