Headline

Data Scientist @ Kenvue | NLP & Generative AI | Python, PyTorch, Transformers | Building Production ML Pipelines

About

As a Data Scientist at Kenvue, I develop and deploy NLP-driven solutions to address business challenges at scale. My current focus is on creating a Bill of Materials (BOM) Cross-Reference pipeline using fine-tuned RoBERTa embeddings and multi-method similarity scoring, achieving 88% match coverage. I also diagnosed and resolved production issues affecting multi-component record matching and implemented a confidence-boosting layer using logistic regression and TF-IDF for improved accuracy. I hold a Bachelor of Technology in Electronics and Communications Engineering from Jain (Deemed-to-be University). My technical expertise spans Python, PyTorch, Transformers, and LangChain, with experience in building and maintaining ML pipelines on Azure ML using CI/CD. Passionate about leveraging AI, NLP, and machine learning to solve real-world problems, I am open to connecting with professionals working on advanced AI/ML applications.

Location

Greater Bengaluru Area

AI stack

claude

langchain

langgraph

openai

githubcopilot

Chroma

Work experience

Generative AI Engineer at Kenvue, Jan 2026 - Present · 10 mos, Bengaluru, - Developed an NLP pipeline to automatically match ERP and TRU (Product 360) Bill of Materials records using fine-tuned RoBERTa embeddings and multi-method similarity scoring, achieving 88% match coverage. - Diagnosed and fixed a production bug causing silent match failures for multi-component records (front/back labels), building a custom rescue-matching function that recovered lost matches without impacting existing logic. - Implemented an ML-based confidence-boosting layer (Logistic Regression + TF-IDF) that improved high-confidence match rates by up to 15%. - Deployed and maintained the pipeline on Azure ML with CI/CD via Jenkins.

Software Engineer at Happiest Minds Technologies, Aug 2023 - Jan 2026 · 2 yrs 6 mos, Bengaluru, Karnataka, India · Hybrid, • Developed and deployed hybrid recommendation engine combining collaborative filtering and content-based algorithms, improving user engagement metrics and click-through rates . • Implemented matrix factorization techniques (SVD, NMF) and deep learning models using TensorFlow to process 2M+ user interactions and generate personalized recommendations. • Engineered comprehensive feature extraction pipeline handling user behaviour patterns, item metadata, and temporal features, improving recommendation accuracy by 25%. • Containerized application using Docker and deployed on AWS EC2 with auto-scaling capabilities. • Implemented CI/CD pipeline using GitHub Actions for automated testing, model validation, and deployment, reducing deployment time.

Engineer Trainee at Happiest Minds Technologies, Aug 2023 - Feb 2024 · 7 mos, Bengaluru, Karnataka, India · Hybrid

Education

Bachelor of Technology - BTech, Electronics and Communications Engineering, Jain (Deemed-to-be University)