Headline

SWE @InMobi | DL enthusiast | DSG | Ex-Intern @Samsung | Electrical Engineering, IIT Roorkee

About

I'm a self-driven ML practitioner at IIT Roorkee with hands-on experience building end-to-end AI systems — from prototyping models in PyTorch and TensorFlow to deploying them with Docker, Kubernetes, and MLflow on cloud platforms like AWS and GCP. My recent work has been focused on large language models, working with Hugging Face Transformers and LangChain to build LLM-powered applications. I enjoy the full lifecycle — not just training a model, but making it production-ready. What I bring to the table: → Strong foundations in Python, deep learning, and NLP → Practical MLOps experience: containerization, CI/CD for ML, model serving → Cloud deployment on AWS/GCP → Hands-on projects in LLMs, Generative AI, and classical ML I'm currently seeking AI/ML or MLOps internship opportunities where I can contribute to real-world systems while learning from experienced teams. Let's connect — I'm always open to conversations about AI, open-source, and building things that ship

Location

Greater Delhi Area

AI stack

claude

Work experience

Software Engineer at InMobi, Jul 2026 - Present · 2 mos, Bengaluru, Karnataka, India

Teaching Assistant at Indian Institute of Technology, Roorkee, Jan 2026 - Apr 2026 · 4 mos, On-site, • Assisted in teaching Applied Machine Learning (EEL 367) to 80+ third-year Electrical Engineering students • Conducted tutorial and doubt-solving sessions on ML concepts and applications • Guided students with assignments and practical implementations

Developer at Samsung R&D Institute India - Bangalore, May 2025 - Jul 2025 · 3 mos, Bengaluru, Karnataka, India, • Enhanced multi-turn tool reasoning in Qwen2.5-3B-Instruct using RLHF (DAPO, GRPO, PPO) in a 3-member team. • Fine-tuned on API-GEN-MIT-5K with custom rewards and GPU-optimized training (DeepSpeed, FSDP), improving throughput by ~10%. • Achieved 65.2% on BFCL-eval and 62.98% on API-BANK, outperforming all other 3B-class models. • Led discussions on GPU optimization and contributed to RLHF pipeline design.

Research Intern at Indian Institute of Technology, Roorkee, May 2024 - Jul 2024 · 3 mos, On-site, • Reviewed convex optimization techniques (Lagrange/Wolfe Duals, KKT) and applied them to SVM variants including LS-SVM, Proximal, GEPSVM, Fuzzy, and Depth SVMs. • Analyzed optimization structures and depth-function-based robustness; documented insights and open problems in a research report.

Core Member at Data Science Group, IIT Roorkee, Apr 2023 - May 2026 · 3 yrs 2 mos, Roorkee, Uttarakhand, India, • DSG aims to work on innovative open source projects in the domain of Machine Learning, Deep Learning and Reinforcement Learning, while also spreading the culture of AI within the campus. • Regularly conducted public lectures, hackathons and competion.

Education

Bachelor of Technology - BTech, Electrical and Electronics Engineering, Indian Institute of Technology, Roorkee

Grade: 12th-94.2%, Maa Bharti School Kota