Inspiration
We've all experienced group projects where everyone thinks they're on the same page, only to realize they had completely different expectations.
Miscommunication, unclear responsibilities, and unspoken concerns often lead to wasted time and unnecessary frustration.
We wanted to build something that addresses these issues before they affect a team's progress.
What it does
Mediator is an AI-powered collaboration platform that helps teams uncover hidden disagreements and establish shared expectations.
Team members independently submit their goals, priorities, deadlines, and responsibilities.
Our AI analyzes different perspectives, detects misalignment, and suggests actionable resolutions.
Rather than making decisions for the team, Mediator helps everyone reach a shared understanding.
How we built it
Designed a multi-agent workflow with three components:
Perspective Analysis: Understands each member's individual input.
Conflict Detection: Identifies contradictory expectations and unclear responsibilities.
Resolution & Synthesis: Generates potential compromises and shared plans.
Incorporated a human approval step so teams maintain control over decisions.
Used Google Antigravity and Figma Make for development and prototyping, with n8n for AI workflow orchestration and GitHub for version control.
Challenges we ran into
Distinguishing actual disagreements from differences in how people phrase their responses.
Finding the right balance between AI-generated suggestions and human decision-making.
Prioritizing our core features within the limited hackathon timeframe.
Accomplishments that we're proud of
Developing an AI-driven approach to identifying hidden team misalignment.
Designing a workflow that transforms individual perspectives into actionable recommendations.
Keeping human collaboration and decision-making at the center of our product.
What we learned
AI can do more than automate tasks; it can help people recognize misunderstandings and communicate more effectively.
Building multi-agent workflows requires balancing context, accuracy, and actionable outputs.
Successful collaboration isn't about avoiding disagreements, but addressing them early.
What's next for Mediator
AI Meeting Shadow: Track decisions, deadlines, and commitments across meetings.
Idea Synthesis: Combine the strongest parts of different team proposals.
Persistent Memory: Remember previous agreements and flag contradictions.
Integrations: Connect with existing project management and communication platforms.
Try it out
Log in or sign up for Devpost to join the conversation.