Srivatsan Ramanujam
Big Data EngineerEngineering Management, ML@Meta
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Engineering Management, ML@Meta
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Seasoned ML and engineering leader with 6+ years in engineering management. Helping organizations build and grow complex customer-facing ML products, with a track record of taking products from conception to launch, successfully navigating production readiness, processes and people challenges. Recognized at growing engineers, managers and building strong teams.
The University of Texas at Austin
MA (MS with Thesis), Computer Science
N/A – Present
PSG College of Technology
Bachelor of Engineering, Information Technology
N/A – Present
Meta
Engineering Management, ML
December 1, 2021 – Present
The Apache Software Foundation
Apache MADlib Project Management Committee Member
July 1, 2017 – March 1, 2023
Salesforce
Software Engineering and ML Leadership, Einstein Platform
August 1, 2016 – December 1, 2021
San Francisco Bay Area
Pivotal Software, Inc.
Principal Data Scientist and Manager
August 1, 2012 – August 1, 2016
Palo Alto, CA
Sony Mobile Communications
Natural Language Scientist (Senior SWE, Software Research)
April 1, 2011 – August 1, 2012
Redwood Shores, CA
Salesforce.com
Member of Technical Staff (SWE)
September 1, 2009 – April 1, 2011
San Francisco
IBM Almaden Research Center, San Jose, CA
Research Intern
May 1, 2008 – August 1, 2008
San Jose, CA
The University of Texas at Austin
Research and Teaching Assitantships
November 1, 2007 – August 1, 2009
Austin, Texas Metropolitan Area
Strand Life Sciences
Senior Software Engineer
May 1, 2006 – July 1, 2007
Bangalore, India
Computer Sciences Corporation
Software Engineer
June 1, 2004 – May 1, 2006
Greater Hyderabad Area
Asset management : excess inventory identification through parts clustering
February 1, 2016 – Present
We built a framework for training hierarchical clustering and connected component analysis models for identifying excess inventory in the products manufactured by a large company. Using PL/Python on Pivotal's MPP databases (Greenplum/HAWQ), we harnessed the machine learning libraries in SciPy and scikit-learn, to build scalable, in-database solutions for asset management.
Text analytics for predicting insurance claims (Insurance)
December 1, 2015 – Present
We built text analytics models to predict insurance claims from structured and unstructured data sources. We harnessed models from libraries like MADlib, scikit-learn and NLTK to build in-database scalable solution for predicting insurance claim amounts on Pivotal's MPP stack (Greenplum/HAWQ).
Subscriber Session Analysis using Topic Models and Clustering
January 1, 2015 – Present
Analyzed subscriber sessions on a class of devices (financial industry) using Topic Models (LDA) and Clustering.
Student Segmentation and Retention Modeling (Education)
April 1, 2014 – September 1, 2015
We built a student segmentation model for a large educational provider and laid the foundation for identifying students at risk of dropping out. We used a variety of source systems including demographics, grade book, discussion boards, course data to build clustering, regression and classification models. The Pivotal Greenplum MPP database was our platform along with MADlib and PL/Python for the machine learning algorithms.
Analysis and Simulation of Frequency Capping on Video Ad Impressions (Digital Media)
March 1, 2014 – Present
We analyzed user consumption behavior of video ads for a large digital media company and modeled the effect of applying frequency caps on ad impression per issuer. To this end we built a simulation pipeline which could analyze the effect of frequency capping on terabyte-scale impression logs and presented insights on the effect of frequency cap on key metrics like ad revenue, burn-rate, forecasted surplus impressions etc. The analysis and simulation was built on MPP (Massively Parallel Platform) on top of Hadoop using a mix of open source tools in Python and C++. Companies can bring in new revenues of the order of a couple of million dollars through our analysis and insights.
Subscriber Segmentation using Messaging System Logs (Telecom)
March 1, 2014 – Present
I analyzed the messaging system logs of a major Telecom provider and used it to build a subscriber segmentation model. The goal of this engagement was to show the value in the messaging logs our customer was already collecting and processing it at scale on the Pivotal Technology stack. The subscriber clusters we uncovered helped our customer in better assessing the growth in their subscriber segments and in capacity planning for the same. We used MADlib and PL/Python (procedural Python) on top of Pivotal's Greenplum MPP database.
Commodity Futures Prediction with Tweets (Social Media)
April 1, 2013 – Present
We used sentiment analysis of Twitter data from GNIP to predict commodity futures for a major agri-business cooperative. Sentiment analysis along with tweet metadata helped us in gathering valuable features in predicting commodity futures. Our sentiment analysis algorithm was semi-supervised and is portable across multiple domains in English. We implemented this at scale on the Pivotal Greenplum MPP database with PL/Python.
Shrinkage Prediction for large Supermarket (Retail)
February 1, 2013 – Present
Shrinkage is the process of lost revenue owing to items getting lost, stolen, wasted or consumed between point of manufacture and point of sale. We used data from along with point-of-sale data to identify groups of stores and products showing similar shrinkage characteristics, this laid the foundation in developing an early warning system to identify products and stores at risk of shrinkage and taking remedial action to minimize shrinkage loss. A ran clustering and co-clustering algorithms at scale across hundreds of thousands of products and hundreds of stores to uncover interesting products and store clusters. The point-of-sale data further were correlated with products worst affected with shrinkage. We used Pivotal's Greenplum MPP database with MADlib to solve this problem.
Topic and Sentiment Analysis of Call Center Logs for Churn Prediction (Telecom)
December 1, 2012 – Present
We blended unstructured data (call center transcriptions) with structured data (demographics, billing history, rate plans, device information, call logs, text statistics) to enhance a churn model's accuracy for a major Telecom company. We used Topic and Sentiment Analysis of call center transcriptions to uncover interesting subscriber behavior that was indicative of Churn. Our blended model delivered significant improvements in Churn prediction for our customer.
Deep Learning Specialization
Coursera
June 24, 2026 – Present
Structuring Machine Learning Projects
Coursera
June 24, 2026 – Present
Sequence Models
Coursera
June 24, 2026 – Present
Neural Networks and Deep Learning
Coursera
June 24, 2026 – Present
Convolutional Neural Networks
Coursera
June 24, 2026 – Present
Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization
Coursera
June 24, 2026 – Present
Cultural Fit Analysis
The candidate's diverse project portfolio across various industries (agri-business, retail, education, insurance, digital media, telecom) and their experience in both large corporations (Meta, Salesforce, Sony, IBM) and open-source communities (Apache MADlib) suggest a high degree of adaptability and a broad perspective. Their leadership roles and contributions to scalable ML infrastructure indicate a strong alignment with a culture that values innovation, impact, and technical excellence. The focus on solving real-world business problems with data science aligns well with a results-oriented environment.
Soft Skills & Operational Fit
The candidate's resume highlights significant leadership experience, managing teams of engineers and ML scientists, and advocating for scalable ML infrastructure. This suggests strong communication, collaboration, and strategic thinking skills. Their involvement in open-source projects and presentations to C-level executives further indicates a proactive and influential operational fit. The numerous project descriptions demonstrate a problem-solving mindset and the ability to translate complex technical work into business value.