Bringing legal contracts onto the blockchain with Ricardian contracts - Smart Contracts which become legally binding.

The problem RicardianAI solves


Welcome to RiccardianAI, an innovative project that transforms legal contracts into Cadence smart contracts for Flow blockchain. Our goal is to streamline the conversion process, leveraging AI and FlowDe IDE for seamless deployment. Here's an overview:


Traditional legal contracts are complex, time-consuming, and prone to human error. Migrating them to Flow blockchain requires manual labor and expertise, hindering adoption for businesses and individuals.

Solution - RiccardianAI

RiccardianAI automates contract conversion by intelligently analyzing clauses, terms, and conditions. It generates secure smart contracts, saving time and minimizing human error. FlowDe IDE simplifies deployment on Flow blockchain.


RiccardianAI offers a user-friendly interface for uploading legal contracts. The intuitive frontend, built with modern web technologies, ensures easy navigation for non-technical users.


The backend utilizes advanced ML algorithms and NLP techniques to analyze contracts. Powered by robust infrastructure, it ensures optimal performance and responsiveness. Built with Python, TensorFlow, and Flask.

Future Development/Improvements


We prioritize speed and plan to optimize and refactor the codebase. Thorough code reviews and best practices will ensure clean, efficient, and well-documented code.


To enhance security, we conduct rigorous audits, implement encryption techniques, and follow secure smart contract development practices


Continuous improvement of the intuitive UI/UX is our commitment. User feedback and usability tests will guide us to refine the interface, streamline workflows, and simplify the conversion process.

Overall Improvements

We strive for excellence and actively seek feedback from users and industry experts.

Challenges I ran into

Getting accurate Cadence smart contracts from a legal contract was challenging and required a lot of training. Creating the correct query and prompt was also challenging but we overcame that by fine tuning the queries to the AI model.

Deploying the contract from the frontend was also challenging, to resolve this we made use of flowDE developer tool api endpoints to achieve the deployment to emulator directly from the browser. (flowDE is a devtool that was also built and submitted for this hackathon)