Headline

Senior Data Scientist | Ex-Spinny | Expertise in Machine Learning, Deep Learning, Computer Vision | Generative AI & Recommendation Systems Specialist | Innovator in Predictive Modelling

About

With a passion for transforming data into actionable insights, I thrive in the dynamic world of data science. My journey has been marked by a commitment to leveraging machine learning and deep learning to solve complex problems and drive business value. At Spinny, I developed a license plate masking system with 99.5% accuracy and crafted predictive models that significantly reduced fraudulent dealer transactions, showcasing my ability to create impactful solutions. My experience spans various domains, including computer vision, generative AI, and recommendation systems. I've successfully implemented change point detection and engineered behavioural-demographic features at Programming.com, enhancing data analysis capabilities. My role as a Data Engineer at SelfDecode honed my skills in preparing and processing whole genome data, further broadening my expertise. I excel in predictive modelling, utilising tools like Python and OpenCV to achieve precise outcomes. My work on content-based recommendation systems has improved engagement metrics significantly. I'm driven by the challenge of applying advanced analytics to real-world scenarios. Eager to bring my skills as a Senior Data Scientist to new challenges, I am open to opportunities where I can contribute to innovative projects that push the boundaries of what's possible with data. Let's connect!

Location

Mumbai

AI stack

antigravity

opencode

claudecode

Work experience

Senior Data Scientist at Programming.com, Mar 2025 - Present · 1 yr 6 mos, Mumbai, Maharashtra, India, * Engineered high-throughput computer vision pipelines for real-time edge deployment, achieving sub-20ms latency with high-fidelity object detection. Optimized YOLO and transformer-based models for production, reducing inference latency by 35% while maintaining strict accuracy thresholds for safety-critical use cases. * Built robust instance segmentation frameworks for privacy-compliant data masking, ensuring zero-failure performance across diverse, large-scale video streams. * Architected parallel dual-extraction systems for field-level validation, significantly reducing false negatives in noisy, real-world environments. Implemented automated data quality checks and edge-case handling for large-scale document processing, mitigating downstream financial and operational risks.

Data Scientist at Spinny, Jan 2022 - Jan 2025 · 3 yrs 1 mo, Gurugram, Haryana, India, • Developed a license plate masking system with 99.5% accuracy using object detection(YOLOv8) and instance segmentation(Segment Anything Model), ensuring privacy compliance and eliminating the requirement of vehicle’s re-photoshoot. • Developed a predictive model to classify fraudulent dealer transactions in vehicle auctions by analyzing current and historical bidding trends, reducing monthly losses by Rs 50 lakh. • Developed a predictive model to estimate the probability of potential customers purchasing a vehicle, optimizing call targeting and reducing calls from 60% to 45% of leads while maintaining a 90% conversion rate. • Developed a content-based recommendation system for the auction team to provide personalized dealer suggestions based on their buying, viewing, and bidding history, achieving 8.5% seen-to-click, 2.5% seen-to-bid, and 29.4% click-to-bid rates. • Developed an OCR system for precise vehicle license plate recognition. Implemented image de-warping using Python and OpenCV to correct skew angles, ensuring accuracy. Leveraged the PP-OCR framework with custom algorithms to achieve a Character Error Rate of 0.5%.

Data Engineer at SelfDecode, Jun 2021 - Jan 2022 · 8 mos, • Prepared, processed, and normalized whole genome data to evaluate imputation quality. • Experimented with enhancing data access performance in a non-cluster-based environment from delta lake. • Created a Dashboard for genetics data analysis and disease trait prediction.

Data Scientist at Taiyō, Feb 2021 - Jun 2021 · 5 mos, 1)Build semantic and syntactical searching system. 2) Maintain embedding and data pipeline from data cleaning, geocoding, to statistic feature generation. 3) Build a system for recommending the most influential economic indicators using a combination of one vs rest modelling, anomaly detection and curating feature to recommend based on the weighted feature importance of each model.

Computer Vision Software Engineer at KritiKal Solutions, Dec 2020 - Feb 2021 · 3 mos, Exploring and experimenting with various ways for table detection, cell detection, and information extraction from invoice documents for OCR performance improvement.

Education

Bachelor of Technology - BTech, Electrical, Electronics and Communications Engineering, Jaypee University of Engineering and Technology