🚀 Software Engineer @ VerifiedTalent 🎓 Integrated B.Tech + M.Tech in Information Technology, IIIT Gwalior ⚡ Focused on backend engineering, databases, AI services, infrastructure, and distributed systems.
Backend-focused Software Developer with experience building production web applications, APIs, automation workflows, LLM services, AI agents and cloud-hosted applications in fast-paced startup environments.
I primarily work with Node.js, TypeScript, Python, PostgreSQL, Redis, MongoDB, Docker, and AWS, with growing experience in FastAPI, RAG, LLMs, AI Agents, LangChain, and LangGraph.
Outside of work, I enjoy building personal projects, exploring new technologies, studying system design and backend engineering patterns, and solving DSA and SQL problems on LeetCode.
Currently exploring:
- LLMs, RAG, AI Agents & Automation
- Database Internals, Caching & Performance
- Microservices & Distributed Systems
- System Design & Scalability
- Infrastructure & DevOps
AI Support Copilot · Python, FastAPI, Gemini, PostgreSQL, SQLAlchemy
An internal support copilot with persistent conversations and a backend-controlled multi-step Gemini tool-calling loop enabling agentic AI behavior. The system functions as an AI agent that can inspect and update support tickets through an allowlisted tool registry with Pydantic-validated arguments, structured failure handling, strict read/write safety boundaries, persisted JSONB tool-call logs, sensitive-field redaction, Pytest coverage, and model-dependent evaluations.
Personal Finance Platform with AI Insights · Next.js, Express.js, PostgreSQL, Redis, BullMQ, Docker, Gemini
A production-aware personal finance application built around transactional consistency and asynchronous processing. The backend combines atomic financial mutations and indexed PostgreSQL analytics with Redis caching and BullMQ workers for non-blocking AI financial reviews, while using structured Gemini outputs, rate limiting, schema validation, and a Docker Compose stack separating the API, worker, PostgreSQL, and Redis services.
Cloud Storage Platform · React, Express.js, MongoDB, Mongoose, Cloudinary
A cloud file-management system that models user storage as an OS-like hierarchical directory tree while separating MongoDB metadata from Cloudinary-hosted binary assets. It supports transactional quota accounting, recursive folder operations, ownership-controlled access, compensating cleanup for cross-system failures, signed downloads and thumbnails, and HLS video delivery.
Languages: Python, TypeScript, JavaScript, SQL, C/C++
Backend & Data: Node.js, Express.js, FastAPI, REST APIs, GraphQL, PostgreSQL, MongoDB, MySQL, Redis, Prisma, SQLAlchemy, Alembic, BullMQ
AI Engineering: LLMs, LangChain, LangGraph, RAG, Tool Calling, AI Agents, LLM Evaluation, Structured Outputs, Embeddings, Vector Search, pgvector, Gemini, OpenAI, Ollama
Cloud & Infrastructure: AWS, Docker, Docker Compose, Nginx, Linux, GitHub Actions, CI/CD
Frontend: React, Next.js, Redux, Zustand, Tailwind CSS
Tools: Git, GitHub, Postman, Pytest, HTTPX, Bash
- A Lightweight Causal Sound Separation Model for Real-Time Hearing Aid Applications — IEEE Sensors Letters
- Channel Estimation in 5G and Beyond Networks Using Deep Learning — IEEE Radioelektronika
- LinkedIn: https://linkedin.com/in/yashveer4
- Email: yashveer.rtc@gmail.com