ConflictSense AI

ConflictSense AI uses Gemini 3 and DigitalOcean Gradient™ to analyze text, voice, and image inputs, detecting early community conflict risks and generating actionable peacebuilding insights.

Comment

ConflictSense AI – Devpost Submission Details Project Name ConflictSense AI Elevator Pitch A multimodal AI platform that detects early signs of community conflict from text, audio, and images, delivering actionable insights to support peacebuilding. Project Story Inspiration Conflicts often begin quietly in communities through rumors, misinformation, or small disputes. Early signs are scattered across text messages, voice notes, and images, making it difficult for local leaders and NGOs to respond effectively. ConflictSense AI was inspired to empower communities with early detection of risks and actionable insights. What it does Analyzes text, voice, and image reports from communities Detects early warning signs of conflict, misinformation, or tension Generates actionable insights and recommendations for leaders and NGOs Displays outputs in a clear, visual dashboard for rapid decision-making How we built it AI Engine: Gemini 3 API for multimodal reasoning Cloud Infrastructure: DigitalOcean Gradient™ for GPU-powered AI inference Backend: Python + FastAPI Frontend: React / Next.js Deployment: Vercel / Railway Database: PostgreSQL / SQLite Version Control: GitHub Challenges we ran into Designing AI prompts that are sensitive and accurate Handling multimodal inputs consistently Presenting complex AI outputs in a simple, actionable dashboard Maintaining ethical and responsible AI practices Accomplishments that we're proud of Built a fully functional multimodal AI platform Integrated Gemini 3 with DigitalOcean Gradient™ for scalable AI inference Created a dashboard usable by non-technical leaders Outputs provide real actionable insights What we learned Multimodal AI provides richer, more accurate insights DigitalOcean Gradient™ accelerates training and inference Clear presentation is key for community adoption Ethical AI design is essential for sensitive applications What's next Real-time alerts for leaders Additional local languages and dialects Partnering with NGOs for multi-community deployment Continuous AI learning via feedback loops Built With Google Gemini 3 API – Multimodal reasoning DigitalOcean Gradient™ – GPU-powered inference Python – Backend and AI logic FastAPI – API server React / Next.js – Dashboard frontend HTML / CSS – UI Vercel / Railway – Hosting PostgreSQL / SQLite – Database NumPy / Pandas – Data processing Try It Out / Demo Links Live Demo: https://peacelens-ai-demo.vercel.app Public Code Repo: https://github.com/yourusername/conflictsense-ai Video Demo Link (Optional: Add YouTube or Vimeo link to your demo video) Example: YouTube Demo Project Media / Image Gallery Hero Image – ConflictSense AI logo + AI network + community theme Dashboard Screenshot – Multimodal input, AI insights, charts Architecture Diagram – Input → Gemini 3 → Gradient™ → Dashboard Multimodal Input Example – Side-by-side text, image, audio waveform Impact Visual – Infographic showing early conflict detection (Make sure all images are JPG/PNG ≤5 MB, 3:2 ratio) Additional Info (For Judges) Submitter Type: Individual Country: Nigeria Public Code Repo: https://github.com/yourusername/conflictsense-ai Functional Demo: https://peacelens-ai-demo.vercel.app Project Status: New DigitalOcean Services Used: Gradient™ AI, GPU Servers, Droplets, Object Storage Resources Available: Yes Experience Feedback: Using DigitalOcean Gradient™ AI was smooth and productive. GPU-powered cloud inference allowed fast multimodal AI processing. Challenges included integrating text, audio, and image inputs, and designing ethical outputs. Suggested improvements: more multimodal workflow templates, advanced monitoring/debugging tools, and team collaboration features. Likelihood of Using DigitalOcean Again: Very likely Primary Status: Full-time Employee / Founder & CEO

Built With

Share this project:

Updates