Contribution to the full lifecycle of AI social product EchoChat.

• Architected a multi-platform social ecosystem across Flutter mobile and Next.js web clients with a microservices backend; integrated Gemini, OpenAI, and Claude for automated Life Story generation, AI-moderated feeds, and image validation/transformation.
• Engineered vector-based embedding systems for user matching and a custom random-walker clustering pipeline across 30k+ geospatial memories, creating a dynamic social map and improving retrieval quality.
• Built Python/Selenium scraping pipelines and agent-driven research workflows to enrich event and social data; deployed serverless infrastructure on Vercel and Supabase with GitHub Actions CI/CD and real-time analytics dashboards.
• Open-sourced the OpenClaw Plugin and EchoMem Skill to sync local markdown memory into EchoMem Cloud and enable retrieval inside OpenClaw workflows.

Contribution to the RESTful APIs for a social app using TypeScript and Next.js.

• Created a user-matching service using TypeScript and Supabase, applying cosine similarity and semantic search to improve recommendation accuracy by 20%.
• Established a performant, scalable, real-time search engine for finding friends using Next.js, Supabase, and Vercel KV.
• Designed and implemented a chat-simulation API leveraging the OpenAI API for dynamic, AI-driven conversations, then optimized backend architecture to reduce response times by 30%.

Contribution to an AI-driven dating app using TypeScript, Next.js, React, and Node.js.

• Utilized TypeScript and modular component design to maintain scalability and support a 35% increase in concurrent users without performance issues.
• Designed and set up comprehensive navigation in Next.js, achieving a 30% increase in user engagement through smoother page transitions.
• Created dynamic camera and profile-selection features and spearheaded Firebase/Google Analytics tracking, boosting personalized interaction by 20% and accelerating new-user registration by 50%.

Cloud database benchmarking and NLP-based resume matching research.

• Adapted YCSB+T to benchmark NoSQL and distributed SQL databases, enabling direct workload comparisons and analyzing performance trade-offs in cloud environments.
• Conducted NLP and transformer-based research for semantic resume matching; improved recruiter efficiency by 25% by aligning resumes to job descriptions and search queries.
• Built visualization outputs for embedding correlations and contributed to open-source platform and educational resource development.

• Managed project documentation across schematic design, design development, and construction documentation phases for a 1.6 million sq. ft. healthcare project.
• Streamlined data workflows between Excel and Revit with Python scripts and collaborated with Mechanical, Electrical, and Plumbing teams for monthly value engineering.