“Manas has been my true partner-in-crime in Search and Browse team at eBay. He is one of the best applied researchers I have had the privilege to work with. What sets Manas apart from the rest is his uncanny ability to identify product opportunities from patterns in vast amounts of user behavior data. He joined our team as an embedded researcher in the early days of Browse team. Within a short span, Manas created a first of its kind data quantification framework to understand search user behavior. We relied heavily on this framework to set our Browse product strategy on Search pages, especially for problem definitions and solution hypotheses. Manas architected and developed several robust, scalable, and contextual recommendation models for each solution hypothesis, and delivered significant business wins. Additionally, this work has resulted in multiple key innovations and patents for eBay. Manas played a key role at every step of the product life cycle -- product ideation, evangelizing solutions across multiple teams, building and deploying the recommendation systems, and iteratively improving the recommendation algorithms via A/B testing. Manas has brought a change in how eBay's green-field projects mature, and has set a template for a successful full stack data scientist in eBay's product development cycle. Any team would be lucky to have him and I'd love to work with him again!”
About
Activity
-
Excited to announce our upcoming workshop: "Multi-Modal Search and Recommendations"! 🚀 As multi-modal approaches become mainstream, leveraging the…
Excited to announce our upcoming workshop: "Multi-Modal Search and Recommendations"! 🚀 As multi-modal approaches become mainstream, leveraging the…
Liked by Manas Somaiya
-
After nearly a decade at Roku, I’ve decided to embark on a new career opportunity. I am grateful to my colleagues and team at Roku for their…
After nearly a decade at Roku, I’ve decided to embark on a new career opportunity. I am grateful to my colleagues and team at Roku for their…
Liked by Manas Somaiya
-
Some of the interesting details allowing a glimpse under the hood of how PYMK recommendations actually work. Happy to collaborate with multiple teams…
Some of the interesting details allowing a glimpse under the hood of how PYMK recommendations actually work. Happy to collaborate with multiple teams…
Liked by Manas Somaiya
Experience & Education
Publications
-
Data-driven co-clustering model of internet usage in large mobile societies
Proceedings of the 13th ACM international conference on Modeling, analysis, and simulation of wireless and mobile systems
-
Mixture models for learning low-dimensional roles in high-dimensional data
Proceedings of the 16th ACM SIGKDD international conference on Knowledge discovery and data mining
Archived data often describe entities that participate in multiple roles. Each of these roles may influence various aspects of the data. For example, a register transaction collected at a retail store may have been initiated by a person who is a woman, a mother, an avid reader, and an action movie fan. Each of these roles can influence various aspects of the customer's purchase: the fact that the customer is a mother may greatly influence the purchase of a toddler-sized pair of pants, but have…
Archived data often describe entities that participate in multiple roles. Each of these roles may influence various aspects of the data. For example, a register transaction collected at a retail store may have been initiated by a person who is a woman, a mother, an avid reader, and an action movie fan. Each of these roles can influence various aspects of the customer's purchase: the fact that the customer is a mother may greatly influence the purchase of a toddler-sized pair of pants, but have no influence on the purchase of an action-adventure novel. The fact that the customer is an action move fan and an avid reader may influence the purchase of the novel, but will have no effect on the purchase of a shirt.
In this paper, we present a generic, Bayesian framework for capturing exactly this situation. In our framework, it is assumed that multiple roles exist, and each data point corresponds to an entity (such as a retail customer, or an email, or a news article) that selects various roles which compete to influence the various attributes associated with the data point. We develop robust, MCMC algorithms for learning the models under the framework.
Patents
Languages
-
English
-
-
Hindi
-
-
Gujarati
-
Organizations
-
Association for Computer Machinery (ACM)
-
- Present -
Institute of Electrical and Electronics Engineers (IEEE)
-
- Present
Recommendations received
8 people have recommended Manas
Join now to viewMore activity by Manas
-
It was an honor to be part of an amazing gathering with press and creators in NY where we shared the AI innovations happening across multiple…
It was an honor to be part of an amazing gathering with press and creators in NY where we shared the AI innovations happening across multiple…
Liked by Manas Somaiya
-
AI should be a catalyst for real progress, but too often we haven't seen that. One of the good things about being at @Microsoft is seeing the sheer…
AI should be a catalyst for real progress, but too often we haven't seen that. One of the good things about being at @Microsoft is seeing the sheer…
Liked by Manas Somaiya
-
A company’s culture is driven by its leaders. At LinkedIn, “Members First” is one of the defining values that drives our culture, meaning we…
A company’s culture is driven by its leaders. At LinkedIn, “Members First” is one of the defining values that drives our culture, meaning we…
Liked by Manas Somaiya
People also viewed
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore More