const freny: Developer = {
  education : "B.Tech CSE @ SVNIT Surat  |  CGPA: 8.60",
  focus     : ["Backend Engineering", "Real-Time Systems", "AI-Integrated Applications"],
  currently : "Building production-grade systems & sharpening system design",
  interests : ["Distributed Systems", "LLM Pipelines", "Competitive Programming"],
  philosophy: "Ship systems that scale — not just code that runs.",
};

Computer Science undergraduate who architects event-driven backends, integrates AI/ML pipelines into full-stack products, and approaches every problem through the lens of scalability and system correctness. From real-time WebSocket infrastructure to RAG-based AI assistants, I build software that works at depth — not just on the surface.

A production-grade sports management platform engineered around real-time data delivery and modular backend architecture.

Engineering Highlights

• Designed a WebSocket layer with Socket.io to support concurrent live match updates with sub-100ms latency across multiple active sessions
• Built a stateless authentication system using JWT + bcrypt with token rotation, ensuring secure session management at scale
• Architected domain-separated backend services — isolated modules for team management, match orchestration, and live event broadcasting
• Integrated an AI coaching recommendation engine that surfaces contextual insights based on match data
• Deployed with a decoupled infra strategy: React frontend on Vercel + Node.js/Express backend on Render

Node.js Express.js Socket.io MongoDB JWT React Vercel Render

A mental wellness platform built around a Retrieval-Augmented Generation (RAG) pipeline for contextually grounded AI responses.

Engineering Highlights

• Implemented a RAG-based conversational AI assistant that retrieves domain-relevant context before generating empathetic, grounded responses
• Designed a Python Flask microservice to host the AI pipeline, with the Node.js backend acting as an orchestration layer between UI and AI service
• Built emotion recognition (DeepFace) and sentiment analysis (VADER) pipelines with end-to-end integration into user sessions
• Structured MongoDB schemas optimized for efficient storage and retrieval of session history and AI interaction logs
• Exposed clean REST APIs with proper error boundaries, rate limiting, and schema validation

Node.js Express.js Python Flask MongoDB RAG Hugging Face DeepFace VADER

A high-integrity online voting platform engineered with biometric authentication, RBAC authorization, and real-time result aggregation.

Engineering Highlights

• Led end-to-end architecture and development — from schema design to deployment configuration
• Implemented Role-Based Access Control (RBAC) with granular permission layers for voters, admins, and auditors
• Integrated biometric-based identity verification as a second authentication factor to prevent fraudulent submissions
• Designed a PostgreSQL + Prisma data layer with normalized schemas, query optimization, and transactional integrity for vote recording
• Built real-time result dashboards with live aggregation pipelines and responsive UI updates on vote events

Node.js Express.js PostgreSQL Prisma JWT RBAC Biometric Auth

Core DSA Domains

• Graph algorithms · Dynamic programming · Greedy & divide-and-conquer
• Binary search · Sliding window · Tree traversals
• Segment trees · Hashing · Bit manipulation

Consistent problem-solver with a focus on time-complexity analysis, optimal space usage, and clean algorithmic thinking — directly applied to backend system design decisions.

"The best systems are not just correct — they are inevitable."

Freny Chauhan · SVNIT Surat · B.Tech CSE '28