Data Science Leader
Hello! I am a New York City based leader of scientists and analysts, and I love working on projects where data is the heart of the problem.
My skills on the softer side include:
- Leadership. I place great value on developing the careers of those who report to me, through learning, growth and mentorship.
- Green field problem solving. I do a lot of my best work when given large unstructured problems and a mandate to come up with solutions for them.
- Construction of taxonomies and vocabularies. I excel at understanding the world through these verbal systems. I do well at breaking down big problems into smaller ones, shaping things up, and finding out all of the contingencies and boundaries of challenging problems. Also a bit of pressure helps too!
- Brainstorming. Connecting parts of systems in novel ways.
On the more technical side:
- Deep neural networks in PyTorch. Transformers have been a huge game changer, especially BERT (and related models)!
- Unsupervised machine learning. Clustering (DBSCAN, k-means)
- Big data tools. Spark, Scala, Postgres and Presto are my favorites.
- Information retrieval algorithms. Minhash and indexing!
- Machine learning and Scientific computing. Python, Pandas, Scikit-learn, numpy, R and so on.
- Networking and computer security
- General programming and linux tools (Java, Make, Git, etc.)
How'd I get here?
Currently I work at Yext, where I am leading Data Science efforts for a new product called AnswersYext Answers helps businesses provide better site search. Our goal is to provide a Google-like search experience for anyone's website. So there's tons of NLP work to be done! On a day to day basis, I'm helping establish new procedures and protocols for use of data at Yext, mentoring, and working closely with product and engineering. In terms of research projects, some of them are some of the greatest hits of information retreival and machine learning. But we're not just using tried and true methods, we are also taking advantage of the huge advances in Deep Learning for NLP, particularly with BERT for question answering, named entity recognition and clustering.
At AppNexus, I led a group of five Data Scientists whose job it was to detect and eliminate fraud.6 Months After Fraud Cleanup, AppNexus Describes Impact On Its Exchange We filtered out botnets, click bots, fraudulent "cash out" sites and fake news. We started in 2015 After filtering for fraud, AppNexus transactions fell by 65 percent. Ultimately, in 2018, AppNexus was acquired by AT&T for around 1.7 billion dollars to become Xandr AT&T in Talks to Acquire AppNexus for About $1.6 Billion.. On a day to day basis, I was in charge of making sure the research group was productive and reacting to needs throughout the company. I was also responsible for working closely with executives, product, marketing, legal and law enforcement. Two International Cybercriminal Rings Dismantled and Eight Defendants Indicted for Causing Tens of Millions of Dollars in Losses in Digital Advertising Fraud.
For my master's degree I worked on a co-supervised project in the laboratories of Dr. Andréa LeBlanc and Dr. Derek Ruths. I'm extraordinarily grateful for the opportunity and the scientific training I received at McGill has shaped me for the rest of my life.
I've always enjoyed novel consulting engagements with various and sundry clients with unique problems. Since 2005 I've taken on various projects for a variety of clients in topics such as security and data analysis to detect ad fraud.
Hobbies and interests
I love cooking, exercise, design, photography, and my environment in Bushwick, Brooklyn. Colorful lights and lasers are my favorite and I love the diversity of NYC.
Work with me
If you think my unique mix of skills might be a good fit for a project you have, send me an email! I'm particularly interested in continuing to grow and develop my soft skills, and I'm quite interested in consulting engagements as well.