Experience

Blueshift

• United States
• Software Development
• 1 - 100 Employee

• Lead Data Scientist (AI/ML)

  • Jul 2019 - Present

  As a part of Blueshift's Data Science team, I work on the core recommendation algorithms that power the product. My work involves researching and implementing various collaborative filtering, content-based filtering algorithms, NLP/LLM based recommendations for candidate generation, ranking, re-ranking and user-feed generation. Actively generating email/site recommendations for Skillshare, Discovery, Lendingtree, USNews, Carparts, Udacity, Zumper, etc. Currently focused on Generative AI… Show more As a part of Blueshift's Data Science team, I work on the core recommendation algorithms that power the product. My work involves researching and implementing various collaborative filtering, content-based filtering algorithms, NLP/LLM based recommendations for candidate generation, ranking, re-ranking and user-feed generation. Actively generating email/site recommendations for Skillshare, Discovery, Lendingtree, USNews, Carparts, Udacity, Zumper, etc. Currently focused on Generative AI using LLMs including model fine-tuning and applied science. Show less As a part of Blueshift's Data Science team, I work on the core recommendation algorithms that power the product. My work involves researching and implementing various collaborative filtering, content-based filtering algorithms, NLP/LLM based recommendations for candidate generation, ranking, re-ranking and user-feed generation. Actively generating email/site recommendations for Skillshare, Discovery, Lendingtree, USNews, Carparts, Udacity, Zumper, etc. Currently focused on Generative AI… Show more As a part of Blueshift's Data Science team, I work on the core recommendation algorithms that power the product. My work involves researching and implementing various collaborative filtering, content-based filtering algorithms, NLP/LLM based recommendations for candidate generation, ranking, re-ranking and user-feed generation. Actively generating email/site recommendations for Skillshare, Discovery, Lendingtree, USNews, Carparts, Udacity, Zumper, etc. Currently focused on Generative AI using LLMs including model fine-tuning and applied science. Show less



Marin Software

• Software Development
• 100 - 200 Employee

• Data Science Engineer

  • May 2018 - Apr 2019

  As a part of Marin’s Data Science team, I have designed ML models to predict performance of Ads, hence enabling intelligent bidding recommendations and budget allocations. The prediction models employ heavy feature engineering on Marin’s colossal historic data store and are based on stacks of Gradient Boosted Decision trees. The models not only improved on the accuracies of previous models, but also increased the coverage by many folds to cover all low/high volume Ads. Furthermore, I… Show more As a part of Marin’s Data Science team, I have designed ML models to predict performance of Ads, hence enabling intelligent bidding recommendations and budget allocations. The prediction models employ heavy feature engineering on Marin’s colossal historic data store and are based on stacks of Gradient Boosted Decision trees. The models not only improved on the accuracies of previous models, but also increased the coverage by many folds to cover all low/high volume Ads. Furthermore, I worked on a Multi-arm Bandit based Budget Allocation Engine that considers Ad Sets as Bayesian Bandits and iteratively allocates budgets to them using Thompson Sampling. I have also worked on implementing K-Prototypes and Decision Tree based clustering algorithms for finding similarity among Ads. At Marin, I used Scala for most ETL jobs and Python for algorithmic proof of concepts. My day consisted of working together with developer teams, scrums, data mining, feature engineering, AI/Ml and reading research papers. Show less As a part of Marin’s Data Science team, I have designed ML models to predict performance of Ads, hence enabling intelligent bidding recommendations and budget allocations. The prediction models employ heavy feature engineering on Marin’s colossal historic data store and are based on stacks of Gradient Boosted Decision trees. The models not only improved on the accuracies of previous models, but also increased the coverage by many folds to cover all low/high volume Ads. Furthermore, I… Show more As a part of Marin’s Data Science team, I have designed ML models to predict performance of Ads, hence enabling intelligent bidding recommendations and budget allocations. The prediction models employ heavy feature engineering on Marin’s colossal historic data store and are based on stacks of Gradient Boosted Decision trees. The models not only improved on the accuracies of previous models, but also increased the coverage by many folds to cover all low/high volume Ads. Furthermore, I worked on a Multi-arm Bandit based Budget Allocation Engine that considers Ad Sets as Bayesian Bandits and iteratively allocates budgets to them using Thompson Sampling. I have also worked on implementing K-Prototypes and Decision Tree based clustering algorithms for finding similarity among Ads. At Marin, I used Scala for most ETL jobs and Python for algorithmic proof of concepts. My day consisted of working together with developer teams, scrums, data mining, feature engineering, AI/Ml and reading research papers. Show less



Sprinklr

• United States
• Software Development
• 700 & Above Employee

• Software Engineer

  • Jun 2015 - Jul 2016

  Spearheading Sprinklr's Benchmarking module, an application that compares brands using their public data over all social channels, I worked on ways to handle big data and applied theoretical concepts to solve challenging problems. I also worked as a backend developer for Sprinklr’s Social Business Index, a product that aims to rank all brands based on their social statistics. As a part of this project, I developed various ranking methodologies and recommendation algorithms. Working at… Show more Spearheading Sprinklr's Benchmarking module, an application that compares brands using their public data over all social channels, I worked on ways to handle big data and applied theoretical concepts to solve challenging problems. I also worked as a backend developer for Sprinklr’s Social Business Index, a product that aims to rank all brands based on their social statistics. As a part of this project, I developed various ranking methodologies and recommendation algorithms. Working at Sprinklr, I was exposed to many new technical skills like ElasticSearch, MongoDB, RESTful web services, Spring framework, Git and web containers like Jetty and Glassfish. I was also able to grasp the intricacies of AWS servers, their configurations and optimisations. https://www.sprinklr.com/benchmarking-insights/ Show less Spearheading Sprinklr's Benchmarking module, an application that compares brands using their public data over all social channels, I worked on ways to handle big data and applied theoretical concepts to solve challenging problems. I also worked as a backend developer for Sprinklr’s Social Business Index, a product that aims to rank all brands based on their social statistics. As a part of this project, I developed various ranking methodologies and recommendation algorithms. Working at… Show more Spearheading Sprinklr's Benchmarking module, an application that compares brands using their public data over all social channels, I worked on ways to handle big data and applied theoretical concepts to solve challenging problems. I also worked as a backend developer for Sprinklr’s Social Business Index, a product that aims to rank all brands based on their social statistics. As a part of this project, I developed various ranking methodologies and recommendation algorithms. Working at Sprinklr, I was exposed to many new technical skills like ElasticSearch, MongoDB, RESTful web services, Spring framework, Git and web containers like Jetty and Glassfish. I was also able to grasp the intricacies of AWS servers, their configurations and optimisations. https://www.sprinklr.com/benchmarking-insights/ Show less



OperaSolutions

• IT Services and IT Consulting
• 100 - 200 Employee

• Software Intern

  • Jan 2015 - Jun 2015

  I was exposed to Natural Language Processing algorithms during my six month long internship in 2015 at Opera Solutions, a big data analytics company, where I developed features for text mining and sentiment analysis. I explored various ETL tools, scripting languages like R, SAS, Python with Pandas and Model-View-Controller architecture. Thanks to several workshops on MongoDB, Neo4J and Hadoop, I learnt the flexibility of big-data and NoSQL DBs. I was exposed to Natural Language Processing algorithms during my six month long internship in 2015 at Opera Solutions, a big data analytics company, where I developed features for text mining and sentiment analysis. I explored various ETL tools, scripting languages like R, SAS, Python with Pandas and Model-View-Controller architecture. Thanks to several workshops on MongoDB, Neo4J and Hadoop, I learnt the flexibility of big-data and NoSQL DBs.



Athabasca University

• Canada
• Higher Education
• 700 & Above Employee

• Mitacs Globalink Internship

  • May 2014 - Aug 2014

  The Mitacs Globalink Research Internship is a highly competitive initiative that attracts the best researchers to Canada. I interned at School of Computing, Athabasca University, Canada and worked on the development of Academics Analytics Tool. The Academic Analytics Tool is a portable software application that allows teachers and learning designers to build queries on data in Learning Management Systems (LMSs). I worked with JavaScript, PHP and… Show more The Mitacs Globalink Research Internship is a highly competitive initiative that attracts the best researchers to Canada. I interned at School of Computing, Athabasca University, Canada and worked on the development of Academics Analytics Tool. The Academic Analytics Tool is a portable software application that allows teachers and learning designers to build queries on data in Learning Management Systems (LMSs). I worked with JavaScript, PHP and PostgreSQL. http://www.academicanalytics.ca/team.php Show less The Mitacs Globalink Research Internship is a highly competitive initiative that attracts the best researchers to Canada. I interned at School of Computing, Athabasca University, Canada and worked on the development of Academics Analytics Tool. The Academic Analytics Tool is a portable software application that allows teachers and learning designers to build queries on data in Learning Management Systems (LMSs). I worked with JavaScript, PHP and… Show more The Mitacs Globalink Research Internship is a highly competitive initiative that attracts the best researchers to Canada. I interned at School of Computing, Athabasca University, Canada and worked on the development of Academics Analytics Tool. The Academic Analytics Tool is a portable software application that allows teachers and learning designers to build queries on data in Learning Management Systems (LMSs). I worked with JavaScript, PHP and PostgreSQL. http://www.academicanalytics.ca/team.php Show less



Indian Institute of Technology, Bombay

• India
• Higher Education
• 700 & Above Employee

• Summer Internship

  • May 2013 - Aug 2013

  I worked at Aakash Development Lab and created web ticketing based customer support portal for Aakash Tablet. Project was mentored by Padma Shri awardee Prof. D.B Phatak and acknowledged by Ministry of Human Resources and Development, India. http://www.it.iitb.ac.in/aakashtechsupport/ I worked at Aakash Development Lab and created web ticketing based customer support portal for Aakash Tablet. Project was mentored by Padma Shri awardee Prof. D.B Phatak and acknowledged by Ministry of Human Resources and Development, India. http://www.it.iitb.ac.in/aakashtechsupport/



Indian Institute of Remote Sensing , Indian Space Research Organization

• 1 - 100 Employee

• Summer Internship

  • May 2012 - Jul 2012

  Working at Photogrammetry Lab under Dr. Vinay Kumar, I worked with raw satellite/sensor data and created an application to organize and analyze it. This project honed my web development skills. Working at Photogrammetry Lab under Dr. Vinay Kumar, I worked with raw satellite/sensor data and created an application to organize and analyze it. This project honed my web development skills.