Vinay Kumar

B.Tech CS graduate specializing in NLP, conversational AI, LLMs, and agentic frameworks with 4+ months of production internship experience. Proficient in Python, HuggingFace Transformers, LangChain, LangGraph, FastAPI, and GCP. Experienced in building RAG-based conversational search systems, LLM fine-tuning (LoRA/PEFT), data ingestion pipelines, and agentic AI workflows.

Deployed 5+ live AI applications with strong emphasis on accuracy, latency optimization, responsible AI, and cross-functional collaboration in Agile environments.

AI / ML Intern at Infosys Springboard (2025-01 – 2025-06)

• Built NLP/OCR document ingestion pipelines with data quality gates via pattern analysis, processing 100+ documents per batch; achieved 60% performance gain via model quantization and adaptive batching (4.2s → 1.7s per document)
• Architected conversational AI-ready REST APIs on GCP (Vertex AI, GCS) with 99.5% uptime, 85%+ test coverage, latency monitoring, and Swagger documentation; ensured responsible AI and compliance standards throughout
• Automated CI/CD pipeline via GitHub Actions reducing deployment cycle from 2 hours to 25 minutes; collaborated with cross-functional teams in Agile sprints to deliver scalable NLP solutions

Software Engineer Intern at TechSaksham (2025-08 – 2025-10)

• Developed FastAPI async microservices handling 200+ concurrent requests; optimized system performance reducing p95 latency 40% (380ms → 228ms) with Redis caching — validated via load testing with 1000+ virtual users
• Delivered independently owned use cases end-to-end: design, development, testing (85%+ coverage), and deployment with Docker CI/CD; documented runbooks for operational reliability
• Optimized PostgreSQL queries using composite indexes and query profiling, improving complex join performance by 50%; performed EDA on large datasets to identify data patterns and quality issues

Bachelor of Technology in Computer Science and Engineering – Government Engineering College, Palamu (2022 – 2026)