Resume

B.S. Computer Science and Economics

Software Engineering Intern — Platform Generative AI · Bethesda, MD

• Built an A2A-enabled multi-agent AI platform from Marriott's Java starter kit, evolving a single-service MCP/RAG prototype into a coordinator-led runtime with specialist agents, routing rules, gRPC communication, streaming APIs, memory, monitoring, and dashboard controls.
• Expanded the platform across 5 backend modules and a React/Vite dashboard, growing the system from 36 to 250 Java classes and from 6 to 105 mapped endpoints across coordinator, analytical, news, mock-agent, and knowledge-layer services.
• Integrated PostgreSQL/pgvector RAG over 147 hotel inventory files using 1024-dimensional embeddings, scoped property retrieval, top-k search, and chunked JSON ingestion to ground agent responses in Marriott property data.
• Added scalability and reliability features including A2A agent discovery, gRPC with HTTP fallback, Redis-backed session memory, configurable routing, multi-level caching, health probes, audit logs, and performance metrics.

• Java
• Spring Boot
• gRPC
• pgvector
• Redis
• React

Research Intern — Data Analytics (Immuno-Oncology) · Washington, DC

• Analyzed high-dimensional flow cytometry data to identify immune cell response patterns across complex datasets.
• Cleaned and validated experimental data with Excel and Python to support reproducible statistical analysis.
• Presented data-driven findings at weekly research symposiums and worked with researchers to interpret results and recommend next-step analyses.

• Python
• Flow cytometry
• Statistics

Agentic AI portfolio and stock analysis platform

• Full-stack investment intelligence platform with autonomous agents.
• Real-time market data, forecasting, backtesting, and portfolio memos.
• Deployed on Vercel and Render with Neon Postgres and resilient model routing.

• React
• TypeScript
• FastAPI
• Postgres
• Plaid
• A2A

Adaptive LLM orchestration and prompt optimization platform

• Routes requests across providers with automatic failover and streamed progress.
• Turns rough ideas into structured prompts while tracking usage, cost, and execution traces.
• Ships with a Chrome MV3 extension and inspectable agent workflows.

• Node.js
• Vercel
• SSE
• Zod
• Chrome MV3

Phone-first object identification and visual learning app

• Captures the clearest frame from a timed camera hold or upload, then returns an object card with confidence, context, use cases, care tips, and shopping links.
• Routes identification through Gemini Vision with a lightweight ONNX MobileNetV2 classifier fallback behind one Next.js API.
• Learns from corrected labels and saves recognized objects into shopping, research, or for-later storyboards.

• Next.js
• TypeScript
• Gemini
• ONNX
• Python
• Vercel

Biomedicines 14(2), 317

• Synthesizes mechanistic evidence for blood pressure variability as a multisystem risk biomarker and evaluates emerging wearable and AI-based approaches that could enable continuous BPV monitoring.