Computer vision for images and video:* brand safety / content moderation for images* brand safety and contextual targeting in long form video content* several models to production at millions of API calls per day scale, 250k hours of long form content processed.* scene detection in long form video - significant R&D improving current SOA* Working with data curation team to make guidelines, build and clean datasets* Working with product teams to frame and define problems for the science team to solve* Lead of computer vision team, built team from 1 scientist to 4.* Creative and relentless problem solver to ensure project success as well as best customer experience.* Innovative best in class visualizations including video rendering for demos and debugging.

Leading the computer vision machine learning practice for Macy's Tech. Solving business problems with R&D and model training.* Lead a project to completion for retail fashion visual similarity at scale. (See previous position experience for details.)* Supervised the development and designed the UX for a web based data annotation tool. Through user sessions developed UX optimizations that increased productivity by 5x for the same amount of time.* Mentored ML Engineers of different levels. Also performed knowledge transfers for a range of audiences from non technical business partners to other ML Engineers and Data Scientists.* Lead R&D for boundary boxing prototypes for fashion editorial images, locating and classifying fashion items in user generated photos.* Lead R&D for using convolutional autoencoders as rich feature extractors with a small representational size.* Performed R&D for fashion image segmentation and masking.* Wrote several Python packages to perform a variety of tasks from efficiently loading, transforming, and augmenting images to generating PDF contact sheets for visual model evaluation. Also wrote packages for production grade API's doing real-time calculations for image similarity.

* Lead project from R&D to production using convolutional neural networks (CNN) in a fashion retail environment to make visually similar recommendations. * Worked with product managers and merchants to develop business requirements. * Evaluated candidate models with business stakeholders and shepherded the project through experimentation and customer facing evaluation. * Wrote production grade python packages for image ETL, training, inference, and live recommendation calculation.* Optimized recommendation python code for 4ms API response time on Dockerized Kubernetes clusters on Google Cloud nodes.* Successfully deployed a scalable and reliable RESTful API that served 500M recommendations through Black Friday traffic loads on the 7th largest (USA, by sales) retail website.

Lead NCE of several Cisco Advanced Services accounts of moderate to large size.Coordinated and represented multiple internal network engineers as the primary client-facing engineer; kept projects on schedule while exceeding client expectations.Delivered quarterly business review presentations with senior client management and engineers.Educated clients on best practices, security audits, network architecture, as well as emerging technologies.Supported 3000+ node network refresh for a large retail client. Provided design and configuration, proof of concept, pilot validation, and support during rollout into production environment.

Troubleshooted live customer issues relating to Cisco IOS as well as performance issues on network devices.Won 3 quarterly customer service awards due to highest average customer survey scores for respective quarters as well as 2 innovation quarterly awards.Created external web docs to document commonly encountered customer issues. Example with 24k views: https://goo.gl/NaaViH