| Project | Core Architecture & Innovation | Stack & Impact |
|---|---|---|
| NEXUS Adversarial Multi-Agent Swarm |
Autonomous 7-agent intelligence swarm (planners, researchers, critics, validators, reconcilers) executing distributed adversarial synthesis with real-time SSE streaming. | LangGraph FastAPI Gemini Next.js 15 Redisβ±οΈ <3 min complete multi-perspective reports with sliding-window rate limiting. |
| RayWise AI Solar Rooftop & Satellite CV |
Satellite computer vision platform for automated rooftop segmentation paired with solar irradiance models for clean energy ROI forecasting. | Next.js 15 TypeScript Convex Clerk Gemini API Three.jsβοΈ 92% segmentation accuracy with 3D solar rendering. |
| WildEcho TaxaMoE Multi-Taxa Bioacoustics |
Multi-taxa bioacoustic classification architecture processing 234 species across 5 biological taxa with probabilistic gating over Perch v2 embeddings. | PyTorch MoE Perch v2 FastAPI ONNX RuntimeπΏ Accepted at CLEF 2026 (0.93957 ROC-AUC). |
| UCRIS Credit Risk Intelligent System |
Hybrid tree-neural stacking model with SHAP explainability for default prediction on 30,000+ banking portfolios. | PyTorch XGBoost SHAP FastAPI NeonDBπ 0.9996 AUC with regulatory compliance reporting. |
| CellTrace APM EV Battery Health & Provenance |
Gemini 2.5 Flash diagnostic agent with 3 Scikit-learn predictive models (SOH, RUL, knee-point) anchored to Polygon blockchain for immutable battery provenance. | FastAPI Scikit-learn Polygon Blockchain Solidity Next.jsπ Real-time telemetry visualizer & failure-risk scoring. |
| ColdChain Automated B2B Lead Enrichment |
Autonomous B2B lead enrichment and personalized outreach automation engine with live WebSocket streaming. | FastAPI Python WebSockets Turso / libSQL Next.jsβ‘ High-throughput automated pipeline. |
| LeaseGuard AI Contract Review & Risk Auditor |
Automated SLA parameter extraction and contract audit engine running RAG-guided risk assessments and counter-offer negotiation strategies. | Next.js 15 FastAPI LangChain ChromaDB NeonDBπ― 50+ contract variables audited at 95% precision. |
| π» Languages |
|
| π§ Machine Learning |
|
| 𧬠Deep Learning |
|
| π Frameworks & Libraries |
|
| ποΈ Databases |
|
| βοΈ Cloud & DevOps |
|
| β‘ Platforms & Services |
|
| π οΈ Tools & Technologies |
|



