Feb 2022 – Present · 4 yrs 1 mo · Bengaluru, Karnataka, India

• Working on Google Assistant.

Sep 2020 – Feb 2022 · 1 yr 5 mos · Bengaluru, Karnataka, India

• Part of International Markets team. Working on building products around pricing.

Oct 2018 – Sep 2020 · 1 yr 11 mos · Bengaluru Area, India

• Leading the development of ML algorithms for Clustr Curation engine. The curation engine is a state of the art information extraction and resolution system which performs entity resolution and mapping at scale. The product is helping SMBs to standardize their product and inventory data helping them make accounting, inventory management and point of sale more efficient and receive aggregate business insights.

Mar 2018 – Sep 2020 · 2 yrs 6 mos · Bengaluru Area, India

• Working on information extraction, information retrieval problems for Small and Medium businesses.

Jul 2016 – Feb 2018 · 1 yr 7 mos · Bangalore

• Building machine learning solutions for enterprise. Currently developing algorithms to develop utilities for large scale networks to detect similar nodes, cluster nodes, anomalies etc. Primarily being developed for Smart City applications.
• Primarily involved in developing data science solutions for Industrial Automation, Smart City, Logistics.

Sep 2014 – Jun 2016 · 1 yr 9 mos · Bangalore

• Worked on developing data driven solutions for automotive, manufacturing and logistics. Worked on building real time, scalable analytics solutions in Predictive Maintenance and Quality, Telematics, In-vehicle Diagnostics using big data technologies.
• Developed a brake monitoring application using in-vehicle data to predict brake state in real time. Patent filed on the technique
• Working on developing real time anomaly detection algorithms for IoT applications including detecting anomalous parts in electronic manufacturing, anomalous devices in a cluster of power tool devices and anomalous vehicles in a fleet of connected vehicles
• Working on analyzing vehicle sensor data to develop indices for driving behavior, fuel efficiency, braking pattern etc. These indices provide insight into vehicle and driver performance. Using Apache Spark for analyzing large amount of sensor data
• Designed KNIME based workflow architecture for key analytic tasks for rapid prototyping of analytic solutions

Oct 2012 – Aug 2014 · 1 yr 10 mos · Bangalore, India

• Currently as part of Lenovo e-commerce, leading strategic initiatives around data driven decision products for marketing and retail applications.
• Developed online visitor segmentation and scoring model. Implemented a regularized logistic regression model to score visitors for propensity to purchase. Model implemented on Lenovo website
• Develeoped multi-channel customer data mining and pattern analysis capabilities. Implemented anomaly detection, segmentation and factor analysis algorithms for deep insights into customer behavior
• Developed interactive dashboards with advance visualization for real time insights
• Performed text mining, sentiment analysis and document categorization

Oct 2010 – Sep 2012 · 1 yr 11 mos · Bangalore, India

• Responsible for applying statistical modeling and machine learning techniques to enhance yield, reduce process variation and improve productivity across semiconductor, display and solar panel manufacturing units.
• Developed and implemented a boosted tree ensemble algorithm with regularization for automatic feature selection in very large feature problems. The algorithm provided ability to identify causes of process excursions and tool failure by identifying the relevant features through a learning model.
• The algorithm is implemented in Applied's Fault Detection and Classification system.
• Developed a parameter search algorithm using particle swarm optimization to perform heuristic search for optimum response curves for predicting machine failure
• Published 2 conference papers on relevance of boosting and tree based ensemble methods for feature selection in semiconductor fault detection problems.

Sep 2007 – Dec 2009 · 2 yrs 3 mos

• Performed research in distributed networked control systems, helping develop models and control strategies for networked systems in industrial automation and smart grids. Worked on developing preventive maintenance strategies using event data from plant floor. Results published at 3 international conferences.

Jul 2005 – Jun 2007 · 1 yr 11 mos

• Ubiquiti Incorporated develops software products for data analytics and text mining. As an analyst I worked on leveraging text mining and data analysis techniques to help Automotive OEMs analyze large sets of repair and warranty data.