Engineering AI That Works in the Real World
I am a Fullstack AI/ML Engineer & Data Analyst with over 3 years of hands-on experience designing, building, and deploying production-grade intelligent systems. My work lives at the intersection of rigorous academic research and real-world software engineering — I do not just study AI, I ship it to production environments that serve real users and drive real decisions.
At GormalOne LLP, Bangalore, I have spearheaded the end-to-end development of multiple flagship AI products — the SSR patented OCR framework that reduced inference latency by 90% through custom multi-threaded parallel processing and NVIDIA GPU runtime optimization, the GauGuru Agentic Multilingual RAG System supporting all 22 scheduled Indian languages via integrated ASR, TTS, and NMT pipelines, and the UIDIB biometric cattle identification system achieving 99% identification accuracy using metric learning and muzzle pattern recognition.
My technical depth spans Large Language Models, Computer Vision, Natural Language Processing, Speech AI, and MLOps. I am experienced in distributed model training on NVIDIA DGX systems, inference optimization with TensorRT and vLLM, containerized microservice deployment via Docker on GPU-enabled Linux servers, and building secure JWT-authenticated REST APIs. I consistently bridge cutting-edge research papers into fully operational production pipelines.
I am currently pursuing my M.Tech in AI/ML (Deep Learning specialization) at BITS Pilani — one of India's most prestigious technical institutions — while working full-time. My undergraduate record at SRM University–AP reflects a 9.08/10 CGPA, demonstrating sustained academic excellence alongside intensive hands-on technical delivery.
Beyond the code, I have authored technical white papers, filed intellectual property, and been recognized as Best Performer and Innovator of the Quarter at GormalOne. I am driven by the conviction that AI's greatest value lies in solving problems that truly matter — food security, livestock management, digital safety, and multilingual access to knowledge across India's diverse population.
Research-First Engineering
Patent filings, white papers, and academic research translated directly into live production systems with measurable impact.
Inference Optimization Expert
TensorRT, Triton, vLLM, NVIDIA DGX — delivering under 3s latency where 60 seconds was the prior production baseline.
Multilingual AI Pioneer
Built full ASR, NMT, and TTS pipelines serving all 22 scheduled Indian languages within a single agentic system.
End-to-End Deployment
From model training on DGX to Docker containers, GPU Linux servers, secure APIs, and Power BI dashboards.
