Inspiration
Navigating regulatory compliance is frustratingly slow and opaque—especially for startups and businesses in emerging markets. Permits can take weeks, and audits are often manual, error-prone, or unactionable. Inspired by this inefficiency, we built RegOps AI Suite to automate the entire regulatory lifecycle using multi-agent intelligence—bridging the gap between complex regulation and actionable guidance.
What it does
RegOps AI Suite is an AI-powered compliance automation platform. It:
Automates permit eligibility, submission, and tracking.
Performs multi-agent audits of documents, logs, and business practices.
Explains compliance risks and resolutions via an interactive chat UI.
Integrates with government sources (via scraping or APIs) to validate status.
Supports regulators with an admin dashboard to review submissions.
All powered by a modular, orchestrated agent system using the Agent Development Kit (ADK) on Google Cloud.
How we built it
Frontend: Built with Next.js and deployed on netlify.
Backend: FastAPI (Python) for REST APIs, following clean code and DDD principles
Database: MongoDB for flexible, scalable data storage
AI Agents:* Modular agent classes for analysis, permit, and audit tasks, leveraging Google Cloud AI APIs (Vertex AI, Document AI)
Email: Mailtrap API for secure, testable transactional email
Infrastructure: Docker Compose for local development, with deployment workflows to Google Cloud VM
Security: Environment variables for secrets, Google Application Credentials for secure cloud access
DevOps: GitHub Actions for CI/CD, automated deployment, and environment provisioning
Challenges we ran into
AI Integration: Adapting our agent architecture to work seamlessly with Google Cloud’s AI APIs required careful design and testing.
Environment Management: Ensuring secure, consistent configuration across local, CI, and cloud environments (especially with Google credentials and MongoDB networking).
Email Deliverability: Integrating Mailtrap for safe, reliable email testing without risking real user data. Dependency Conflicts: Navigating Python package version mismatches (e.g., Motor/PyMongo, Pydantic v2+).
Cross-Platform Development: Handling differences between Windows, WSL, and cloud environments, especially for Docker and MongoDB.
Accomplishments that we're proud of
Built a robust, modular backend that’s easy to extend and maintain
Seamlessly integrated Google Cloud AI services for real-world compliance automation
Achieved secure, automated deployment to Google Cloud with minimal manual steps
Implemented a clean, testable email workflow using Mailtrap
Maintained strong security and best practices throughout (env vars, secrets, health checks)
Created a developer-friendly codebase with clear documentation and onboarding
What we learned
- The importance of clean architecture for rapid iteration and scaling in AI projects
- How to leverage Google Cloud AI Development Kit for fast, secure, and scalable AI integration
- Best practices for managing secrets, credentials, and environment variables in cloud-native apps
- The value of automated CI/CD and infrastructure-as-code for hackathon velocity
- How to design AI agent interfaces that are both powerful and easy to extend
What's next for RegOps AI Suite
Deeper AI Integration: Expand agent capabilities with more Google Cloud AI APIs (e.g., natural language, vision, and translation)
User Experience: Build a modern frontend for compliance teams and regulators
Compliance Templates: Add industry-specific workflows and document templates
Audit & Explainability: Enhance audit trails and AI explainability for regulatory trust
Marketplace: Enable third-party agent and workflow plugins
Production Readiness: Harden security, add monitoring, and prepare for enterprise deployment on Google Cloud
Log in or sign up for Devpost to join the conversation.