ODDS is a continuous learning community for Data Science leaders

First data science hire of the company and built the team from the ground up. Focus of the team was on recommender systems, ranking, and forecasting. We partner with Product, Product Engineering, Finance, and Marketing to build production ML systems and internal capabilities.Team building and strategic leadership:• Developed staffing plans, interview process, and career matrix • Coach and mentor both data scientists and analysts• Identify company needs. Develop intra- and inter-team processes and create roadmaps to meet business priorities. • Lead cross functional communication and education to the larger organization on data science uses, data product development life cycle, and experimentation. Integrated data science as a key function across business verticals.Technical:• Acted as senior individual contributor and launched many algorithmic data products, primarily in Product-centric domains, to surface and recommend relevant products to improve customers' shopping experience with Grove and drive greater engagement• Stood up critical ML infrastructure for production systems (Airflow, Sagemaker, Feature Store, CI/CD, Docker)• Led the development and best practices on end-to-end ML system design patterns for integrating algorithms as a service and for a/b testing algorithms, including model training, deployment, updating, monitoring, and tracking• Provided key insights to the broader organization through exploratory analysis that have changed company OKRs and reshaped company strategy

As part of the Inventory Optimization Team, I've led a cross functional effort in launching and scaling an internal capability that powers how we stock our key merchandise and manage our inventory efficiently, through all phases of the product life cycle. This includes • Conceptualize, build, and continually improve on inventory forecasting models to predict future inventory state, recommend inventory needs, and optimize inventory buys• Develop and maintain ETL pipelines and APIs to deploy models that integrate with the business• Architected the system design from data models to user work flows, including acting as PM to develop a user facing application that allows merchandisers to interact with algorithms• Partnered cross functionally to define strategic vision, drive alignment, establish processes, and implement change managementPreviously I worked on helping launch the Mens business line. Responsibilities included performing exploratory analysis and deep dives on business outcomes in relation to clients and merchandise, guiding the design of experiments, and building dashboards to monitor key business metrics.

• Perform exploratory data analysis and machine learning on medical claims and physician meta data to quantify physician quality, derive insights about the clinical process, and match high quality physicians to patients.• Work cross functionally with Medical, Care, Engineering, and Product Teams to deliver end-to-end data-driven product changes, including prototyping data models, simulating impact, coordinating/validating data integration and production implementation, and gathering performance feedback.• Build and maintain data infrastructure and ETL jobs internal to the Analytics Team.• Onboard new data scientists and mentor junior members of the team.• Tools: Python data science stack (Pandas/Matplotlib/Scikit-Learn/SQLAlchemy), MySQL, Google BigQuery, web scraping (Selenium/BeautifulSoup), D3, Google APIs, AWS, Amazon MTurk, Airflow.

• Developed novel, automated denoising algorithm for magnetoencephalography (MEG) measurements to study neural responses in human visual cortex. • Developed computational models to capture statistics of high-dimensional time series data (MEG & EEG), using PCA, GLM, and Fourier analysis.• Parallelized model fitting and computation using high performance computing clusters.

• Research used signal processing frameworks and computational models to explain sensory and cognitive processes and neural responses. • Designed, implemented, and analyzed experiments involving physiological measurements at different spatial and temporal scales. Developed custom software for data collection, analyses, modeling, and visualization.• Techniques included: regression, classification, Monte Carlo methods, bootstrapping, Bayesian statistics, constrained optimization, cross-validated model comparison, hypothesis testing, dimensionality reduction, generalized linear models, time-frequency analysis, deconvolution.

• Developed neuroscience course and lab material for undergraduates. Led laboratories, recitations, exam reviews, and one-on-one tutoring sessions. Graded homework and exams. • Courses taught: Natural Science II: Brain & Behavior; Behavioral & Integrative Neuroscience.

Implemented and performed numerical simulations of biophysical neural circuits and studied their nonlinear dynamics.

Researched preference-based decision making. Implemented and executed human psychophysical experiments and developed software for macaque psychophysics.

Implemented and evaluated noise-robust automatic speech recognition algorithms.