Ran cybersecurity data science. Organization of 17 people contained research, operations, engineering, and analytics sub-teams.Responsible for all R&D underlying the dozens of models at the heart of the Cylance suite ofcybersecurity productsPartnered with executives, product teams, and engineers across company to foster machinelearning/big data/analytics best practices within BlackBerryLed algorithm research and implementation, model operationalization, and production evaluationEntirely rewrote team’s deep learning code base, speeding up training by a factor of ten andincreasing accuracy of primary classification models by more than a factor of fifteenTech stack: Python, Pytorch, Petastorm, AWS S3/EC2/Athena, Jenkins, BitBucket, Spark, Docker

Ran the Kensho Los Angeles office; oversaw all hiring effortsLed the Kensho Extraction project to automate information retrieval from tens of millions offinancial documentsCoordinated the Kensho machine learning journal group

Oversaw all data science efforts at OpenX; team included four local and six remoteemployeesInterfaced with executive/senior management to advise business and strategic directions and determine optimal application of data science amid broader analytics initiativesOrganized and led bi-weekly sync of all analytics teams at OpenXLed the team that entirely rewrote the OpenX engineering career ladder (used for performance appraisal for all engineering ICs)Invented novel models to address programmatic bid request filtering, inventory forecasting, and buyer bidding insights

Led all technological efforts of the data science team; mandate included rapid prototyping, assessment, and implementation of deep learning and other advanced machine learning solutions to address business needsPartnered with product, strategy, engineering, and BI teams on initiatives including business development, customer segmentation, customer insights, user behavior modeling, competitive intelligence, speed, system resource usage, and forecastingInvented and managed the deployment of suite of predictive algorithms providing an immediate growth in revenue in primary business areaPioneered a novel request filtering technology enabling instantaneous assessment of relative worth of traffic; algorithm frees significant portion of system capacity across all OpenX servers with negligible revenue impact

Created and implemented numerous algorithms using Python, R, Matlab, C++, and bash to perform and optimize feature extraction and selection, predictive modeling, data munging, and other analysis of large high-dimensional datasetsFounded company's internal machine learning group; served as chief technical advisor on wide array of machine learning effortsSpearheaded company's deep learning research and achieved state-of-the-art performance on multiple datasets through application of deep networksActed as Project Manager/Principal Investigator on multiple SBIRs

Principal Investigator on a Phase II SBIR and a Phase II STTR.Invented a beamforming algorithm employing parametric and heuristic methods to enable a multistatic radar configuration to rapidly resolve a few hundred co-interfering moving objects. This algorithm was used as the technological basis for three separate Phase II SBIR/STTRs.