Headline

Systems Engineer @ TCS | GenAI & SLM Fine-Tuning Specialist | Python | QLoRA | PyTorch

About

I’m a final-year Computer Science student with a strong passion for AI and real-world problem-solving. I thrive on building practical solutions — whether it involves experimenting, iterating, or piecing together concepts from the ground up. From training models to testing new workflows and diving into code, I find the most joy in learning by doing. I approach projects with a hands-on mindset — always researching, testing, failing, and refining until things click. My work blends logic with creativity, and I'm constantly looking to sharpen my skills in system design, AI, and innovative thinking. I’m passionate about contributing to meaningful projects that push my growth and leave an impact. I'm open to opportunities in AI and software development where I can apply my strengths, stay curious, and keep building things I’m excited about.

Location

Navi Mumbai

Work experience

System Engineer at Tata Consultancy Services, Jun 2026 - Present · 2 mos, Thane

Technology Consulting Intern at EY, Jan 2026 - May 2026 · 5 mos, Sāki Nāka, Enterprise Architecture: Analyzed multi-tier enterprise systems across the Persistence, Logic, and Service layers to evaluate backend data flows. Data Modeling: Designed structured data models and managed database architecture using Core Data Services (CDS) View Entities. Systems Integration: Utilized data interface protocols (OData) to maintain data integrity and optimize performance across distributed networks.

Machine Learning Intern at Quant Matrix AI, Jun 2025 - Nov 2025 · 6 mos, Hybrid, Engineered scalable generative AI pipelines and optimized open-source Small Language Models (SLMs) for domain-specific enterprise intelligence. Focused heavily on fine-tuning methodologies, dataset alignment, and infrastructure optimization for deployment in resource-constrained production environments. ​Key Contributions ​SLM Fine-Tuning & Parameter Efficiency: Specialised in fine-tuning state-of-the-art open-source architectures (including DeepSeek-R1, Llama 3.2, Phi-3, and Llama-7B) using LoRA and QLoRA parameter-efficient techniques. ​Compute & Memory Optimization: Leveraged the Unsloth framework to implement 4-bit model quantization, successfully accelerating training throughput by 2\times while drastically shrinking system memory overhead. ​Data Engineering & Alignment: Built robust custom NLP data preprocessing pipelines and tokenization schemes to manage domain-specific alignment, optimizing models for targeted institutional use-cases. ​Performance Benchmarking: Collaborated on design frameworks for precise model evaluation metrics, validating accuracy, perplexity, and inference limits to maintain stability under tight compute boundaries.

Education

Bachelor of Technology - BTech, Computer Science, Ramrao Adik Institute of Technology