Developed a prototype of an advanced sensor system for anonymized patient monitoring, crucial for privacy-centric healthcare environments. The system effectively determines the number of people in diverse room settings using non-intrusive sensors and no cameras for patient privacy.Configured NVIDIA Jetson Nano as a standalone access point, eliminating the need for existing network infrastructure and enhancing system versatility in various healthcare settings.Utilized ESP32 modules for motion and temperature data acquisition, ensuring the capture of essential environmental metrics while maintaining patient privacy.Trained a neural network on thermal sensor data, to accurately identify the number of occupants in a room, an effective approach to people counting without visual surveillance.

Fortify is a provider of High Performance Additive Manufacturing technology. The software organization develops CAD/CAM tools for preparing 3D models, software and firmware that runs the hardware, and cloud based data collection tools for performance analysis, optimization, and maintenance. Fortify Compass is a 3D Modeling and CAD application for laying out and supporting models for prints. Led the team to develop this application, designed and architected the code and some of the key numerical algorithms. Flux Host is a tablet-based application that runs on the printer and communicates with the state machine that runs the various components in the firmware. I co-designed the system and project managed its development. Fortify Flux Developer is implemented an IoT style infrastructure that allows for printers to upload information during prints. The information is organized in a Relational Database. The web interface, data access microservice and cloud database run on Google Cloud Platform. This enables real time monitoring of internal and customer prints, post print analysis, statistical analysis, machine maintenance monitoring, and process engineering for development of new materials.Provisional Patents: 63/172,654, 63/173,324, 63/174,519

Joined HMH full time to build, as part of a small team, a production Data Pipeline with Kafka, Spark, PostgreSQL and Spring Boot based micro-services running on Amazon Web Services. This pipeline serves large data processing handling peak loads of 10K events per second and averaging 1K events per second. Student performance data is collected and processed for efficient delivery of district, school, class and student reports.

HMH makes web based software for helping K-12 students learn and improve their reading and writing skills. I worked as a front end developer creating data driven web pages oriented for teachers and administrators to understand student performance. This required Javascript, Angular and D3 skills. Additionally I coordinated closely with a third party that provided our data platform.

Developed technical web pages for a variety of markets utilizing JavaScript, D3, CSS, HTML and no-SQL databases. Instanced back-end servers on AWS running Linux to host databases.Developed a complete solution for web-based assessment of severity of retinopathy of prematurity by clinicians for a research project. The system, built on an AWS Linux Server, utilized a CouchDB back-end and provided a set of front-end tools for clinicians to rank, rate and compare images. Created visualization of the results using D3.js for the presentation of ranked order, graphical models of user consistency and inter-expert agreement. This system is used support machine learning algorithms for automated grading of image severity. To facilitate the use of CNC machines by hobbyists, created a set of web pages for image creation (2D parametric CAD) and toolpathing (CAM). Enabled a generic way for users to describe 2.5D geometry by interpreting a grey-scale image as a height map. Using Meijster's algorithm for calculating a Euclidean distance transform, generated offset paths. Rendered the paths back in the (HTML) scene and enabled download of Gcode.

Part of very small team developing PreForm - a CAD / CAM application for 3D printing.

Built web based tool path generation and simulation package 123D Fab for generating G-Code to drive CNC machines. Toolpathing library was written in C++ and communicated as a plugin with Javascript that handled the user interface.

Developed a geometric model analysis tool for integrating changes from a set of parametric-history free edits back into a history based model. US Patent 20120084060.

Helped facilitate the integration of the Product Lifecycle Management tool Compass into the Autodesk Product Line (Productstream).

Responsible for hiring 10 developers and a development manager in Singapore to own the integration of Autodesk Vault with AutoCAD. The role transitioned from interviewing, hiring and managing to technical leadership.

Managed various teams working on Autodesk Inventor including the Part Modeling Group and the API team.

Responsible for Pro/Engineer’s Assembly compute pipeline. Enhanced the dependency sorting algorithms to enable complex parametric cross-assembly relationships to compute in order. Developed an integration of RASNA’s motion simulation software with Pro/Engineer.