Cornell University

Netaji Subhas Institute of Technology

Staff Machine Learning Engineer at Twitter

LinkedIn

Senior Engineering Manager, Machine Learning

Leading teams working on feed ranking, notifications ranking, knowledge graph and safety machine learning that impact top line metrics such as weekly active users, engagement and retention. 
Teams using modern ML techniques including deep neural networks, transformers etc for content understanding, user understanding and engagement prediction models.

Nextdoor

Head Consumer Machine Learning

Leading teams working on feed ranking and trust and safety

Senior Engineering Manager - Machine Learning

Founder and lead of the Conversations Quality team. Developed and operationalized a 3+ year vision for 0->1 ML product to show personalized and non-toxic tweet replies to users for the 2 most visited surfaces in the Twitter app.
- Built the first version of reply ranking from scratch, got buy-in from leadership to build, mentor, and grow a team,
delivering +200M User Active Minutes/day, +3M quality Daily Active Users, +50% positive engagement rate (CTR) on replies. Increased ad-slot inventory by $300M/year via ML investments on tweet replies that reduced user reports for negative content by 45%.
- Bootstrapped a team of 7 ML engineers. Hired 5 MLE’s across senior and staff levels, conducted quarterly performance management and goal setting, coached the new team and mentored 10+ MLEs. I promoted 5 people across different levels in junior and experienced ML engineers. 
- Driving the technical strategy and roadmap across the ML modeling, infrastructure, and data science teams for Conversations to enable scalable, reliable and cost-effective engineering. This resulted in a data pipeline that handles ~20B tweet impressions/day, backend systems that handled feature fetching and server side inference of ~20M Tweets/Sec with 5K+ ML features. The team used modern ML tooling(PyTorch, Tensorflow) and deep learning architectures – MLP, Multi-task learning, Embedding Tables.
- Developed heavy ranking labels for reciprocity, quality replies and health remediations to solve challenges unique to conversations. Developed a light ranking solution that can handle 4x bursts of traffic due to viral tweets with little impact to user experience. This led to filing of 6 patents (3 granted).
- Collaborated with XFN product and data science partners for product vision & planning. I established data driven frameworks for prioritization and culture of rigorous A/B experimentation (10+ every quarter). This led to my team meeting/exceeding OKRs for 6+ quarters.

Twitter

Engineering leader - Staff ML Engineer

Skills: Team Management · Data Products · Organization Skills · Machine Learning Algorithms · Engineering Management · Data Engineering · Engineering · Data Science · Long-term Vision · Communication · Technical Leadership · Analytics · Artificial Intelligence (AI) · Technical Architecture · Predictive Analytics · Computer Science · Iterative · Data Analytics · People Management · Deep Learning

Senior machine learning engineer

using machine learning to rank the the timeline and surface relevant content

Machine Learning Engineer

• Anomaly detection in communities around hashtags by using seasonal hybrid ESD. Used a slack bot to make it easier for people to label system output and evaluate the performance. Link: http://bit.ly/RJ-BeyondTrendingTopics
• Implemented existing research on named entity recognition to perform tweet segmentation into phrases after literature review.

betaworks

Data Science Intern

• Proactive Security Design Reviews for products on multiple platforms such as Reader Mobile Android, LiveCycle, Primetime and PlayPanel.
• Conceptualised and developed web application to track 3rd party libraries which was used by 200+ internal product teams.
• Developed dashboard to visualise security training status within the company. Link: http://bit.ly/asset-dashboard

Adobe

Security Research Engineer

• Engineered the client side of the cloud version of SocialCalc on Palm Pre for engineering and marketing department in enterprises and knowledge institutions. 
• Developed a port for SocialCalc, a JavaScript based spreadsheet application to Palm WebOS for Palm Pre PDA’s.

Rohit Jain