Software Engineer | Data & AI Systems Builder
Turning messy data, vague ideas, and real-world problems into clean, scalable systems. ⚙️

I’m a Computer Engineering undergrad who enjoys working end-to-end — from problem framing and algorithms to production-grade systems and polished UX.

My edge sits at the intersection of:

• Strong fundamentals (DSA, systems thinking)
• Full-stack engineering (modern web, APIs, real-time systems)
• Applied AI & data (LLMs, analytics, evaluation, automation)

I’m less interested in academic demos and more obsessed with things that actually work in the real world — tools people can use, systems that scale, and decisions backed by data.

I’ve built developer tools, AI-powered platforms, analytics systems for businesses, and research-backed ML pipelines — across startups, internships, and national-level hackathons.

• Java (primary) — DSA, algorithms, problem-solving (480+ LeetCode)
• JavaScript / TypeScript — full-stack development
• Python — data analysis, ML pipelines, backend services
• SQL

• React, Next.js, Node.js, Express
• FastAPI, Redux Toolkit, Streamlit
• Prisma, WebSockets, REST APIs

• Pandas, NumPy, Matplotlib
• LangChain (LCEL), RAGAS
• OCR, PyMuPDF
• LLMs: Gemini, CodeLlama, Qwen, Ollama-based models

• Git, Docker
• AWS (EC2, Lambda, S3)
• MongoDB, PostgreSQL, MySQL
• Redis, RabbitMQ
• Postman, Vercel, Cloudinary

• Building production-grade AI systems (not demos)
• Designing developer-facing tools with strong UX
• Evaluating and benchmarking LLMs for real-world constraints
• Deepening systems + backend fundamentals
• Writing cleaner, more maintainable, scalable code

Winner — National Hackathon (EdTech Track, NIT Surathkal)
AI-driven platform that identifies skill gaps, generates personalized learning paths, integrates career tools (resume, jobs, interviews), and plugs directly into GitHub, VS Code, and Figma. Built as a real product, not a pitch deck.

First Runner-Up — State-Level Hackathon
Real-time disaster tracking using GDACS, geospatial climate analytics, multilingual alerts, and an A*-based route optimization system for logistics and carbon-efficient planning.

A modern API testing platform with tabs, autosave, hotkeys, Monaco JSON editor, WebSockets, and structured request logging. Built with serious attention to state management and developer experience.

Analyzed sales data across 300+ shops, built Streamlit dashboards, and delivered weekly insights that compared month-over-month profit, product performance, and store efficiency.

Built a natural language interface over MySQL using LangChain, enabling managers to query internal data conversationally instead of navigating dashboards.

• I enjoy badminton 🏸 and competitive problem-solving
• I care about clarity — in code, design, and communication
• I prefer simple architectures that age well
• I’m allergic to over-engineering and buzzword-driven systems

• GitHub: github.com/Vallen328
• LinkedIn: linkedin.com/in/vallen-dsouza
• LeetCode: leetcode.com/u/sc1n4ZJ45W
• Email: vallen2608@gmail.com

Build things that survive contact with reality.
— Vallen