Nov 2022 – Jun 2025 · 2 yrs 7 mos · Bengaluru, Karnataka, India · Remote

• Principal Data Scientist | Line Financial | Oct 2022 - June 2025
• Led AI innovation at Line Financial, developing a cutting-edge multi-agent ecosystem that delivered personalized financial guidance to customers. Architected specialized financial agents with advanced memory systems and implemented dynamic model selection strategies that optimized performance - reducing latency while maintaining high accuracy standards.
• Key achievements:
• Built risk management solutions for the microlending platform, enhancing decision-making capabilities
• Developed forecasting systems that significantly improved collection prediction accuracy
• Created a transaction intelligence platform that reduced data enrichment costs and streamlined operations
• Designed scalable AI infrastructure supporting multiple specialized agents with sophisticated memory architectures
• This role allowed me to combine cutting-edge AI research with practical financial applications, driving measurable business impact through innovative technology solutions
• Skills: AI architecture, multi-agent systems, machine learning, NLP, time series forecasting, agile leadership

Nov 2019 – Oct 2022 · 2 yrs 11 mos · banglore

• Senior Data Scientist | Huawei Music | Nov 2019 - Oct 2022
• Architected and implemented comprehensive music recommendation systems serving users across 20+ countries, significantly improving user engagement and content discovery metrics.
• Designed and deployed multi-layered data models (base, common, scenario-specific) to efficiently capture user preferences and content popularity across diverse regional markets
• Built a robust A/B testing framework with variant identification enabling near real-time performance evaluation and optimization
• Implemented advanced machine learning solutions including Multi-Armed Bandit algorithms for solving playlist cold-start problems, ALS matrix factorization and Neural Collaborative Filtering for personalized recommendations
• Developed sophisticated content matching using Word2Vec for co-occurrence similarity and multi-faceted similarity algorithms
• Created a unified system architecture capable of handling diverse regional preferences while maintaining consistent performance standards
• Engineered audio feature extraction and deep neural networks for improved genre classification and content discovery.

May 2017 – Nov 2019 · 2 yrs 6 mos · Noida, Uttar Pradesh, India

• Data Scientist | Gaana.com | May 2017 - Oct 2019
• Spearheaded innovative audio recognition technologies and data-driven user engagement solutions for one of India's leading music streaming platforms.
• Developed India's first ambient music recognition system using advanced audio fingerprinting technology, enabling users to identify songs playing nearby.
• Built a comprehensive audio fingerprinting platform with dual-tier architecture for efficient music deduplication, implementing offset search on low-level fingerprints to prevent popularity fragmentation
• Created sophisticated user behavioral analysis models for predicting retention, churn, and optimizing targeted advertising
• Engineered an XGBoost classification model to accurately infer user demographics (age group and gender) from listening patterns, enhancing ad targeting capabilities
• Implemented a dynamic audio advertisement generation system that improved conversion rates
• Enhanced the "Made for You" recommendation feature by incorporating ALS matrix factorization and rule-based co-occurrence similarity algorithms, significantly improving user engagement metrics
• Developed and deployed an automated lyrics acquisition system through web crawling, expanding the platform's content offerings.