Square’s Product ML team is responsible for developing impactful ML and AI solutions across all Square products. I lead this amazing group of data and ML engineers who are experts in the domains of conversational AI, multimodal AI, applied ML and ML infrastructure.

Square is Block’s original starting point and a global leader in technology that supports business of all kinds. Square’s goal is to provide state of the art software and hardware that supports our vision of economic empowerment for all. My team sits within Square’s Seller Experience organization which facilitates and measures the unified customer journey, intelligently empowering customers through Square's ecosystem of products. Making sustained progress allows us to significantly enhance Square’s go-to-market strategy and help the company to grow and gain competitive advantage.My team’s primary focus is building and supporting machine learning infrastructure for various ML teams across Seller Exp. These teams work on models that support core functions such as product recommendations, customer churn, ML-powered sales and account management and marketing ML just to name a few. We partner with all these teams by creating a frictionless and modern machine learning development lifecycle. Our solutions encompass the entire lifecycle from data gathering/management, feature engineering, automated model training/retraining, model management/access, model serving and monitoring.

Team Lead for the AI Signal Extraction team. Established a new team that was responsible for integrating applied ML solutions into Bloomberg's enterprise trading solutionsFor the sell-side software solution (TOMS), we worked on ML driven sell-side order ranking. The task here was to improve the trader workflow by reprioritizing trading routes and requests for quotes. Details of our work was also presented at Columbia's Machine Learning in Finance workshopFor the buy-side (AIM), we worked on adding ML-driven portfolio manager decision support. The task here was to focus the attention of portfolio managers on those positions we believed required action based on market conditions and past portfolio manager actions

Team Lead for the Artificial Intelligence - Market Analytics team. News moves the market. At Bloomberg, we strove to not only provide the headline/story as fast as possible but to annotate that narrative with impactful, NLP-driven insight. Various case studies have shown how Twitter is usually a place of breaking news. My team owned the central backend service that consumed a real-time stream of tweets and applied various NLP techniques such as language/spam filtering and topic classification. All of this with latency constraints that required sub-10ms processing times on average. In another project, it became apparent that tackling information overload for our clients was key. With the daily flood of news stories from 100,000s of sources, our clients were having trouble keeping up with what was happening when searching for news. We developed a news clustering and summarization system that powered the NSTM function on the Bloomberg Terminal as a novel solution to this issue. This system was featured at a top-level NLP conference (ACL 2020) for which my team was incredibly proud to be participants in. See our paper on NSTM in the links below.

Team Lead responsible for AIM's Shared Services and Audit team. The team was responsible for creating and designing service-oriented APIs for trading data acquisition. The team was also responsible for the real-time auditing backend of AIM which audited all transactions happening throughout the system.

I was responsible for the design and development of order creation services within Bloomberg's AIM department. AIM is Bloomberg's buy-side order and portfolio management software solution.

Software Engineer in the Interest Rate Derivatives team where I was mostly responsible for infrastructure around our interest rate volatility products.