Shenggang Ding

Software Engineer Intern

• Built synthetic data generation workflows producing 10+ datasets, reducing real-world data dependency by 30%
• Deployed and monitored 10+ forecasting models achieving ~90% accuracy for user well-being signals
• Developed interactive analytics dashboard, reducing feature-prioritization turnaround by 20%

Algorithm Developer Intern

• Designed 2 end-to-end ML pipelines for equity forecasting, improving Sharpe ratios by 15%
• Created hybrid genetic-algorithm + ML framework expanding factor universe from 600 to 800+
• Optimized 120 high-frequency alpha factors, contributing to 5% increase in portfolio returns

PersonaWeave 🤖

Built an LLM-driven framework for computational social science with psychologically-grounded agent personas. Implemented Big Five & Dark Triad trait conditioning (87% correlation) with FAISS-based memory supporting 3,000+ concurrent agents with <100ms retrieval. Featured "silent arbiter" policy maintaining balanced participation (Gini < 0.3).

Kanting 🎥

Full-stack platform for real-time lecture recording, transcription, and summarization. Implemented OpenCV+SSIM slide-transition detection for keyframe extraction. Built browser-based streaming with FFmpeg compilation to MP4, AWS S3 storage, and live broadcast capabilities.

AnchorNotes 📝

Android app with location & time-based smart reminders. Spring Boot REST backend with PostgreSQL and JWT auth. Features geofence-triggered notifications, Markdown editing, media attachments, and one-tap PDF export.

TravelSnap ✈️

Travel journaling app consolidating photo journals, geotagged notes, and collaborative itinerary planning. Shareable timeline/map experience emphasizing fast capture and social sharing.