Curtis Mitchell Email & Phone Number

San Francisco, Ca, Us

Washington, Dc, Us

Within the US Census Bureau, I work on the xD team (https://www.xd.gov/) as an emerging technology fellow. The emerging technology fellowship is a program intended to experiment with and apply cutting-edge technology solutions to improve government services within the Census Bureau and with domestic and international government partners.Some of the projects I'm working on include:- Partnering with the National Institutes of Standards and Technology (NIST) to create documentation and interactive testing environments for privacy-enhancing technologies (PETs) and privacy-preserving machine learning.- Working with the United Nations Privacy-Enhancing Technology Lab (UN PET Lab), other national statistical agencies, and the open-source group OpenMined to develop an application architecture that will allow private, joint computations on international trade data across multiple remote datasets (see my presentation on this work at PEPR 2024: https://youtu.be/69_RGhIJwMU)- Creating online and open-source tools to help document machine learning models used within the federal government: https://bias.xd.gov/resources/model-card-generator/tool/- Maintaining the xD team's various web applications and cloud infrastructure.

New York, Ny, Us

I was a core team member volunteering in my free time to build open-source software that blends machine learning with privacy-enhancing technologies (PETs) to enable discoveries using privately-held datasets.I worked with several teams within the OpenMined organization, including with the software engineering and documentation teams on the Syft platform (https://github.com/OpenMined/PySyft) that enables machine learning and data science code to execute on remote datasets. Previously I worked on the web application that hosts OpenMined's free online courses (https://courses.openmined.org/). I was also selected for the first round of the Padawan mentorship program where I paired up with OpenMined engineers to take deep technical dives into various aspects of OpenMined's software projects (https://blog.openmined.org/work-on-ais-most-exciting-frontier-no-phd-required/), and I have served as a mentor for subsequent rounds of the program as well as an author and reviewer of the program's written material and software examples.

Moffett Field, California, Us

I worked as a software engineer and scrum master on the Ames Simulation Laboratories (SimLabs) team employed through the contracting firm SYMVIONICS (https://www.symvionics.com). My work focused on full-stack applications for the Air Traffic Management-eXploration (ATM-X) research project, a software platform for exploring how to best integrate drones and air taxis into existing air traffic control infrastructure: https://www.nasa.gov/aeroresearch/programs/aosp/atm-xMy software engineering responsibilities included new feature development, code testing and refactoring, writing documentation, and pull-request reviews. My scrum master responsibilities included sprint planning, leading daily scrum standups, backlog ticket grooming, and working directly with our internal and external project partners and stakeholders to develop feature requirements and project expectations.Some of the languages, libraries, and tools I used included TypeScript, React, MobX, Node.js, NestJS, GraphQL, Nx, MongoDB, PostgreSQL, Python, Flask, and Docker.See the attached UTM project video for an overview of one of the ATM-X projects, and an acknowledgement of my contributions to this project in a research paper on our platform tests in North Dakota.

San Francisco, California, Us

Springboard Machine Learning Engineering Career Track.The program consists of 400+ hours of hands-on course material, 1:1 industry expert mentor oversight, and completion of 1 in-depth capstone project. Skills acquired cover all levels of the machine learning stack, including data wrangling at scale, deep learning, and building and deploying large-scale AI systems.My capstone project was a Japanese-English translation application, using a dataset of 2.8 million Japanese-English sentence pairs to train a neural network and using the resulting model to create a full-stack web application.Project site: https://curt-mitch.com/project/jp-en-translatorProject walkthrough: https://curt-mitch.com/post/jp-en-translator-walkthrough

San Francisco, California, Us

Mode provides a web platform for bringing together SQL, Python, and R along with custom visualization and reporting tools to enable data scientists and analysts to be more productive.Achievements:- primary engineer on a full-stack feature used daily by customers to capture screenshots of Mode reports- organized and lead a JavaScript workshop for members of the customer support engineering team- greatly expanded unit testing for Mode's TypeScript codebase, especially for our state-management system- added interactive tooltips and data summaries to native charting system- regularly responsible for deploying and managing releases and acting as a first responder for possible production bugs and outagesThe languages and libraries used in Mode's primary web application include TypeScript, Angular, RxJS, Redux, Sass, HighCharts, and Webpack on top of a Ruby on Rails backend. Python was also used in parts of Mode's notebooks feature.

Ayasdi's platform and web applications use a mixture of machine learning and topological data analysis (TDA) to find new insights and relationships in data for customers from multiple industries and business sectors.Achievements:- Responsible for migrating Ayasdi's primary statistical view from an older Backbone.js UI to one utilizing React and Redux for improved performance of tables and visualizations- Wrote frontend components for a major feature allowing users to select nodes in one visual graph and have those selections represented simultaneously in other shown graphs as well as the statistical summaries of the data contained in the selected nodes- Regularly presented and demonstrated new features at company all-hands- Wrote multiple Python-based Selenium scripts to test UI features and workflowsThe languages and libraries used on Ayasdi's web applications included React, Redux, Backbone.js, Zepto/jQuery, d3.js, WebGL, and Python

Fremont, Ca, Us

Bluenose specialized in developing tools that help SaaS customer success teams better manage their existing accounts and reduce customer churn.My work included:- Built new and redesigned UI components including a SQL query builder and custom date selector for retrieving data- Interfaced directly with product managers, designers, and other engineers to spec out and test new featuresMy primary tools included AngularJS, D3.js, jQuery, and Less (CSS pre-processor) used on top of a Grails backend.

Milan, It

An international consulting firm offering services in numerous industries. I was part of their alternative energy and energy efficiency program evaluation team that worked with various utility companies and government agencies.My work included:- Performed field data collection for building code evaluation- Created Excel dashboards using VBA to model energy usage across multiple climate zones throughout the United States- Wrote data analysis programs in SAS

Milan, It

An international consulting firm offering services in numerous industries. I was part of their alternative energy and energy efficiency program evaluation team that worked with various utility companies and government agencies.My responsibilities included:- Performed statistical analysis on utility customers’ energy usage- Wrote data mining SAS scripts to find relevant subpopulations of utility customers- Acted as a field engineer to install data loggers for collecting data on energy usage and solar panel efficiency

Machine Learning Engineering Career Track

Web Development, Javascript

Bachelor'S Degree, Physics (Bs), Mathematics (Ba)