ModPilot AI

• 

  Main moderation triage dashboard showing High Risk, Medium Risk, Low Risk, and Needs Review queues with real-time moderation prioritization.

• 

  Expanded High Risk moderation analysis showing detected signal, confidence scoring, AI recommendation engine, and moderator action controls.

• 

  Moderator removal confirmation workflow with threat explanation, moderation reasoning, and protected action validation before post removal.

• 

  Reinstatement workflow displaying moderation history, removed content records, threat scoring, and moderator recovery controls.

• 

  Medium Risk moderation review demonstrating warning recommendations, contextual signal analysis, and human-guided moderation decisions.

Inspiration

Moderators deal with large volumes of posts every day, but most moderation tools still rely heavily on manual scanning and fragmented workflows. I wanted to create a system that visually prioritizes moderation decisions, explains why content was flagged, and helps moderators react faster without replacing human judgment.

ModPilot AI was designed as an intelligent moderation triage assistant for Reddit communities using Devvit.

What it does

ModPilot AI classifies posts into four moderation risk categories:

• High Risk
• Medium
• Low Risk
• Needs Review

The system analyzes moderation signals such as:

• external links
• promotional language
• account age
• low karma indicators
• financial language
• missing flair
• urgency phrases

Each post receives:

• a confidence score
• detected moderation signals
• threat visualization
• AI recommendation
• moderator action tools

Moderators can:

• remove posts
• approve posts
• send warnings
• move posts into review
• reinstate removed posts

The interface was designed to feel fast, modern, and readable during high moderation activity.

How I built it

ModPilot AI was built using:

• Devvit
• Reddit Developer Platform
• TypeScript
• React-style Devvit components
• Custom moderation scoring logic
• Google Gemini AI (for plain-language summaries on medium/high risk posts)
• Custom UI animations and moderation HUD design

The moderation system uses layered signal scoring to classify posts by risk level while maintaining moderator control over final actions.

Challenges I ran into

One of the biggest challenges was designing moderation workflows that felt intuitive while still looking visually polished and information dense. A significant challenge during final development was implementing proper server-side security, including moderator authentication, protected API endpoints, and secure API key management through Devvit's encrypted settings system. Balancing:

• moderation clarity
• signal visibility
• UI responsiveness
• action flow
• visual hierarchy

was difficult during development.

Another challenge was making moderation states behave consistently when actions like remove, approve, and reinstate occurred across different queue states.

Accomplishments that I'm proud of

I am proud that ModPilot AI evolved into a complete moderation experience instead of only a basic dashboard.

Key accomplishments include:

• fully interactive moderation flow
• dynamic risk classification
• custom moderation HUD styling
• AI-assisted moderation recommendations
• visual threat analysis system
• moderator action tracking
• secure moderator authentication and endpoint protection
• premium cybersecurity-inspired interface design

What I learned

This project helped me better understand:

• Devvit application structure
• moderation workflow design
• state management
• UI/UX prioritization
• moderation signal analysis
• AI-assisted tooling concepts

I also learned how important moderation clarity and speed are for real communities.

What's next for ModPilot AI

Future plans include:

• moderator analytics and moderation trend reporting
• community-specific rule tuning and scoring profiles
• persistent moderation history and audit tracking
• collaborative moderation workflows
• improved onboarding for new moderators
• expanded AI-assisted moderation explanations
• additional moderator feedback and usability testing

Built With

• devvit
• google-gemini
• javascript
• react
• reddit-developer-platform
• typescript