Ben Sloane Email & Phone Number

San Francisco, Ca, Us

Machine Learning @ Square - Payment Platform

New York, Ny, Us

As the lead data scientist for both our internal workflow tool and our internal assignment engine I built several end to end data products to help key stakeholders make decisions around new product roll-outs, gain insight into inefficiencies in our origination process, optimally allocate staff, and predict customer cycle times.Some of my projects include:• Implementing and deploying a machine learning model to predict loan origination and underwriting durations. This included ETL to create features for the model, simulation software to test multiple strategies/models, deploying the model into production, and designing and managing the experiment to track the model performance. This model helped to reduce fees associated missed deadlines while maintaining downstream conversion, and as a result we found significant cost savings. • Analysis into customer activities to identify bottlenecks during the loan origination process. Utilized unsupervised learning to understand which activities tend to covary with each other with respect to cycle time. This analysis resulted in process changes for how to structure the order of customer activities.• Building a Markov Chain model to understand how customers transition and behave during the underwriting process which will be used to determine cycle time of event transitions and which event is most likely to occur next.

Cary, Nc, Us

As a Data Scientist at SAS I am focused on developing and deploying data & machine learning pipelines. Most recently I have been focused on using deep learning algorithms across many vision and text applications. Some of my projects include:• Development of an end-to-end machine learning pipeline for real-time facial recognition. Utilized YOLO algorithm for face detection, FaceNet to generate facial embedding, and KNN to determine identity from known embeddings database. • Developed and deployed a real-time sentiment analysis model to determine customer sentiment from product review data. Model is a logistic regression which utilized pre-trained GloVe embeddings for text vectorization as features. • Designed a computer vision pipeline to better understand traffic congestion in partnership with city DOT. This includes a CNN for Object Detection and utilized DeepSORT for Object Tracking. With these two models I was able to identify, track, and count individual cars. Using this pipeline, I am looking at various metrics to quantify congestion across multiple intersections. • Provided consultation for a large financial institution which entailed building machine learning models to predict customer churn probability and architecting their machine learning platform to support 10+ person data science team.• Contributor to SAS Python package for deep learning (DLPy), currently focusing on the following features: - TensorBoard integration for tracking model training performance, assessment metrics, and model architecture graphs. - Model pruning API to reduce footprint and computation time during inference

Earth Prime is a start up focusing on sustainability and affordable low maintenance gardening devices. As an engineering intern I am responsible for the creation, implementation, and testing of all components for their first vertical hydroponic system: The iGardenX. My roles include: •Writing software to interface multiple microcontrollers with various sensors (pH, temperature, etc.) •Designing nutrient feeder system •Designing waterproof casing to house circuitry and microcontrollers •Data analysis - Used to determine the accuracy of data received from sensors regarding plant growth

New Orleans, La, Us

•Utilize experimental nonlinear optics to study various phenomena in the field of quantum information science•Interfacing oscilloscope, spectrum analyzer, and other lab equipment to automate data acquisition

Clarks Summit, Pa, Us

• Designed electrical power, lighting, fire alarm, and demolition plans on AutoCAD • Analyzed cost and efficiency of overcurrent protection devices • Consulted with various foremen and construction workers to specify design goals for construction projects • Analyzed National Electric Code to understand and maintain electrical safety requirements

Bachelor’S Degree, Bse-Engineering Physics

Post-Baccalaureate, Computer Science