Apr 2024 – Present · 2 yrs 2 mos · Navi Mumbai, Maharashtra, India · On-site

• Search Query NER: Built transformer-based NER model, improving artist detection (~2% CTR, ~1% STR gain).
• Deepsearch: Conducted research on feature engineering and model improvements for Learning-to-Rank; integrated MLflow for experiment tracking and analysis
• Phonetic Matching: Developed phonetic indexing pipelines for variant matching & new release visibility.
• Query Transliteration: Improved Indic Transliteration service (8+ languages) for multilingual search.

Jul 2022 – Apr 2024 · 1 yr 9 mos · Navi Mumbai, Maharashtra, India · On-site

• Search WebSocket APIs: Led migration to tab-based infinite scroll UI with WebSocket APIs and event-driven workers, reducing latency by 30% via smart caching and lazy loading.
• ML & Data Pipelines: Designed and maintained Airflow DAGs for ML training, A/B testing, Solr ingestion, popularity scoring, user log processing, and feature extraction (Python, PySpark).
• Solr Optimization: Tuned Solr query handlers and field relevance; deployed Near Cache for frequent queries, cutting latency by 9%.

Jul 2021 – Jun 2022 · 11 mos · Bengaluru, Karnataka, India · Remote

• Worked on research related to optimal placement of Integrated circuits(ICs) on processor chip using Graph Neural Networks(GNN)
• Built a processor board UI with customized features using python tkinter that enables placing ICs on board based on data points generated from GNN model which helped in better visualization of results

Jan 2019 – Apr 2019 · 3 mos · Bhubaneswar, Orissa, India

• Worked on a problem statement focused on prediction of closing prices of Indian Stock Indices using a machine learning model under the guidance of Prof. Dr. Ganapathi Panda. Built an Exteme Learning Model(ELM) for faster and accurate prediction of closing prices and published a research paper on the same.

May 2018 – Jul 2018 · 2 mos · Gandhinagar, Gujarat, India

• Worked on a problem statement that centered on categorizing diabetic retinopathy at five severity levels. Under the direction of Prof. Dr. Jignesh Bhatt, I conducted research on cutting-edge deep learning models for image processing, created an All-CNN model for the classification of diabetic retinopathy, and published a research paper on the same