The Manifold Science Cloud enables researchers to publish high-impact research faster. Researchers can use the platform to collect, organize, and analyze multimodal biomedical data to accelerate and improve precision health. As Software Architect at Manifold, I directly support the success of analyst and researcher user profiles using Science Cloud. I design and implement software product solutions which enable them to perform mission-critical research more effectively.

Sr. Manager, Engineering developing critical product features for the Manifold Science Cloud. Researchers can use the platform to collect, organize, and analyze multimodal biomedical data to accelerate research and improve precision health.· As Technical Lead, guided the development of product features in support of analysts and researchers using Science Cloud · Coached growth of direct reports through mentorship via 1-1s, informal feedback, technical guidance, and performance reviews· Architected and implemented one-click self-serve access to secure and scalable compute environments in the cloud, with specialized analytics software pre-installed· Introduced functional and data access control to Science Cloud research projects

Engineering Manager leading a team of 6 engineers spanning multiple disciplines (machine learning, data engineering, devops, MLops) focused on delivering a self-service Machine Learning Platform· Coached growth of direct reports through mentorship via 1-1s, informal feedback, technical guidance, and performance reviews· Developed quarterly engineering plans and collaborated with Technical Lead and Product Manager to deliver ML Platform in feature-prioritized manner· Led team of engineers to implement self-service feature engineering workloads for data scientists to independently deploy data pipelines to production, orchestrated with Airflow· Designed and led development of a scalable ML Feature Store using Snowflake and Feast to deliver a highly performant feature storage and serving system, with data governance best practices built-in by default

Technical Lead managing a team of 4 engineers spanning multiple disciplines (machine learning, data engineering, devops, MLops) · Led team through design reviews and guided system architecture decisions to deliver production machine learning platform· Led team of engineers to implement and deploy parallel serverless data pipelines to production, orchestrated with AWS Stepfunctions· Led development of Spark pipeline for feature engineering at scale of billions of records, using AWS Elastic MapReduce (EMR)· Designed and led development of MLOps orchestration system for batch model inference built in Terraform and utilized for serving inferences to production applications · Designed and built Terraform module for synchronizing AWS Glue Database metadata across AWS accounts and enabling cross-account access to a production feature store· Implemented machine learning solutions using a variety of techniques such as anomaly detection for time-series data

Highly cross-functional role at AI engineering company focused on discovering and developing data products. My role spanned across machine learning and data engineering. Machine Learning — · Implemented machine learning solutions using a variety of techniques such as computer vision and anomaly detection· Co-developed a custom machine learning experimentation framework using Airflow, Kubernetes, and MLFlow· Published two peer-reviewed machine learning papers in collaboration with a digital therapeutics company on predicting the likelihood of patients achieving their healthcare goals in the program· Presented at AnacondaCon 2019 and meetups on the use of Dask for out-of-core and parallel processing for machine learning. Link to co-presentation of Dask at AnacondaCon 2019 is attached.Data Engineering — · Developed robust data pipelines using Airflow, Kubernetes, and Python to ETL data from disparate sources and formats into a unified data store· Worked with scalable data stores such as Amazon Redshift, Azure SQL Data Warehouse, and Elasticsearch

Project: Computer Vision based Person Tracking in a Multi-Camera System· Built a state of the art deep learning based multi-camera person tracking system· Explored recent literature to implement a Python/TensorFlow computer vision pipeline by employing YOLOv2, DeepSort, Kalman Filtering, and custom clustering algorithms· Proof of concept is currently being bench-marked using real world video dataProject: Predictive Maintenance for Connected Oilfield· Developed time series model predicting the probability of equipment failure based on sensor data· Experimented with a number of features, models (Random Forest, Gradient Boosting) and hyper-parameter optimization· Improved model performance and increased understanding of business problem by identifying critical predictive features

· Investigated specific brain regions by integrating computational modeling and neuroimaging techniques· Led global team to develop and execute a ten-session functional magnetic resonance imaging (fMRI) study· Analyzed high-dimensional longitudinal neuroimaging data using Statistical Parametric Mapping (SPM) in MATLAB

· Trained and directed computer camp staff aimed at empowering youth in STEM· Oversaw all administrative tasks, daily logistics, student and staff health and safety· Initial hire as Instructor (2 summers), promoted to Assistant Director (1 summer), and Director (2 summers)