Agent & Agent

Inspiration

Morgan & Morgan's high-volume practice needs intelligent automation. We built an AI Orchestrator that processes messy multi-channel inputs (emails, medical records, case information), identifies actionable tasks, and routes them to specialized AI agents - from Client Communication drafters to Evidence Analyzers - creating a swipe-based decision interface that learns from lawyer feedback.

What it does

Dashboard where legal teams click on cases and receive AI-generated suggestions from specialized agents. Swipe right to accept recommendations, left to pass - making high-volume case management fast and intuitive.

How we built it

React + Next.js frontend, Fast API backend, MongoDB database, Google Gemini AI, with an SDK-based agent orchestrator coordinating specialized AI workers.

Challenges we ran into

• Integrating MongoDB with FastAPI endpoints
• Orchestrating multiple AI agents with proper task routing
• Building intuitive swipe UI/UX for legal workflows
• First-time SDK agent orchestrator implementation
• Trying to integrate Snowflake for text-to-speech

Accomplishments that we're proud of

• Clean MongoDB integration
• Polished UI with Framer Motion animations
• Strong team communication and spirit
• Functional multi-agent orchestration system

What we learned

• FastAPI database-to-frontend architecture
• Effective AI tool prompting and orchestration
• MongoDB + Snowflake integration
• SDK agent coordination
• Asking for help accelerates progress
• Modern web dev with Next.js and Framer Motion

What's next for Agent & Agent

• User authentication and case upload
• Reminders on mobile device for upcoming deadlines
• Expanded agent specializations

Built With

• fastapi
• figma
• framermotion
• mongodb
• next.js
• react
• snowflakeapi