A major obstacle in the blockchain field is obtaining complicated processing and data richness while preserving verifiability. Blockchains lack natural support for large-scale parallel processing, especially high-performance AI workloads, but they are tuned for deterministic execution and decentralized consensus.
From March 7–9, student developers and researchers from UC Berkeley, UWaterloo, and ETH Zurich participated in the Flare x Google Cloud hackathon, organized in partnership with Blockchain at Berkeley, to investigate how blockchains can facilitate AI-driven computation without compromising verifiability.
The participants made use of Google Cloud Confidential Space, a secure enclave that generates cryptographic attestations and offers hardware-enforced confidentiality via the use of Trusted Execution Environments (TEEs). The Flare blockchain allows for the verification of these attestations, guaranteeing that calculations were carried out accurately and without outside intervention. Flare Time Series Oracle (FTSO), Flare Data Connector (FDC), and Flare’s embedded data protocols are ideally designed to provide data provenance and integrity verification for artificial intelligence.
With more than 460 participants in four tracks, the hackathon produced verified AI solutions ranging from Consensus Learning to AI x DeFi (DeFAI). Confidential Space instances with virtual Trusted Platform Module (vTPM) attestations operating on the AMD Secure Encrypted Virtualization (SEV) platform were used in all projects. These attestations functioned as cryptographic evidence that calculations were carried out in a secure and impenetrable setting.
Rene Kolga, Product Manager, Google Cloud stated:
”We’re thrilled to see developers using our Confidential Space secure enclaves to build trusted AI solutions. Projects like 2DeFi and the Flare AI Kit’s launch are exciting steps towards a future where AI and blockchain ensure data integrity and computational assurance.”
2DeFi: Seamless TradFi to DeFi Transition with AI
Two University of Waterloo students named Alex and Hitarth created the notable project 2DeFi. Their project placed second in the RAG Knowledge track and first in the AI x DeFi (DeFAI) track.
2DeFi used Google Gemini AI to address the problem of connecting TradFi customers to DeFi on Flare. Users were given the option to submit screenshots of their Robinhood portfolios, which Gemini 2.0 then examined to determine how risk-tolerant they were. After that, an automated DeFi strategy that included staking and liquidity provision on Flare was mapped to the AI-generated risk profile.
The complexity of conventional Web3 onboarding was removed by 2DeFi’s integration of embedded wallets, which allowed Google login-based wallet setup on Flare and further improved user experience and onboarding.
Flare AI Kit: Empowering Verifiable AI Agents on Blockchain
The formal launch of Flare AI Kit, an open-source SDK for creating verified AI agents on Flare with Google Cloud Confidential Space, coincided with the hackathon. Flare AI Kit was used in the hackathon entries, demonstrating its potential as a foundation platform for blockchain applications driven by AI. Flare AI Kit offers developers a strong security architecture and safe, attestable AI execution environments, making it easier to integrate AI while maintaining verifiability, provenance, and AI safety.
It started with the Flare x Google Cloud Hackathon. Flare AI Kit will grow into a potent, production-ready platform for verifiable AI applications as developers and academics continue to investigate the nexus between blockchain and artificial intelligence. The objective is to make Flare AI Kit a turnkey solution and release it on the Google Cloud Marketplace so that developers and businesses alike can use it as a stand-alone, production-ready solution with continuous enhancements to its capabilities, usability, and integrations.
The Flare AI Kit seeks to revolutionize the development, implementation, and trustworthiness of verifiable AI applications globally by integrating Confidential Space TEE, AI, and the Flare blockchain. Are you prepared to begin developing on Flare? For comprehensive resources and documentation, check out the Flare Developer Hub.