Inspiration
Legal research is expensive, slow, and locked behind subscriptions, making it unsuitable for autonomous AI agents. We wanted to build a pay-per-query legal intelligence platform powered by AI and instant blockchain payments.
What it does
JaaS (Jurisprudence-as-a-Service) is an AI-powered legal intelligence engine that answers jurisprudential queries on demand. Using x402, Arc, and USDC nanopayments, it enables autonomous agents to access legal knowledge without subscriptions.
How we built it
- Multi-agent architecture with Gemini 3 Pro as the orchestrator
- Featherless Qwen2.5-3B for legal doctrine extraction
- x402 HTTP Payment Protocol for pay-per-query execution
- Arc blockchain and Circle Developer-Controlled Wallets for USDC settlement
- REST APIs with real-time query processing and citation generation
Challenges we ran into
- Designing efficient routing between multiple AI models.
- Delivering accurate legal reasoning with reliable citations.
- Making sub-cent payments economically viable.
- Optimizing latency while maintaining high-quality responses.
Accomplishments that we're proud of
- Built a fully autonomous legal AI service with pay-per-query pricing.
- Integrated x402 and Arc for instant USDC nanopayments.
- Achieved high gross margins through efficient AI routing.
- Successfully executed 50+ consecutive on-chain legal queries with 100% success.
What we learned
We learned that combining specialized AI agents with blockchain-based micropayments creates a scalable, cost-effective model for delivering legal intelligence without traditional SaaS subscriptions.
What's next for JaaS — Jurisprudence-as-a-Service
- Expand support to multiple legal jurisdictions.
- Integrate official legal databases and case repositories.
- Add multilingual legal reasoning and document analysis.
- Launch developer APIs and enterprise integrations.
- Enable AI-to-AI legal services for autonomous business workflows.
Built With
- ai/ml-api
- arc-(circle-l1)
- circle-web3-services-(developer-controlled-wallets)
- featherless-ai-(qwen-2.5)
- google-gemini-3-pro
Log in or sign up for Devpost to join the conversation.