Josh Magee Email & Phone Number

United States

Hi! I’m Winston Wolf. I solve problems. (I am under a strict NDA and cannot discuss this solutions publicly, here).

501c3 charity which sends superheroes to visit sick children and their families. All proceeds go directly to provide children and families with gifts, such as shirts, hats, books, etc.Our mission is to entertain ill children by appearing to them as superheroes, empowering them to continue their fight. We aim to inspire, strengthen hope during their struggle, helping them recognize the hero within.

Montreal, Quebec, Ca

Emerging products and rapid development. Drove delivery of a “premium location assessment” POC in 2-weeks; developed half a dozen descriptive and time forecasting models for neighborhood demographics shift and gentrification potential.Added attribution and explainability measures to 2+ internal models (housing AVMs), enabling clients to trace model outputs to specific combinations of input features, bolstering trust in automated pricing mechanisms.

Mountain View, California, Us

Instructor for updated course Advanced Data Visualization with Matplotlib and Seaborn for the Data Analyst Nano degree.Audited Introductory Python course and for update.

Palo Alto, California, Us

Weather Risk MLE; client facing product development with insurers, insurance inspectors, and regulatory team. Regulatory reviews and documentation. I was also with the band (saxophone)!Spearheaded Wildfire Intelligence v2, a suite of Machine Learning models to predict likelihood of wildfire occurrence and claim probability, increasing claim and loss ratio segmentation by 200%.Re-optimized Wildfire v3 Risk model for a client improvement challenge, which had unique data provider requirements. Submitted final model in 8 business days with 5% improved success segmentation from original v3.Developed Hail FastTrack claims model in a single week. This first-of-class model provided the probability of roof damage after hail exposure with 3.5% False Detection Rate and 15% Recall.Collaborated with the regulatory team, 3+ insurance clients, and regulators to develop model filings for California Department of Insurance compliance, focusing on state-of-the-art explainable AI/ML wildfire modeling techniques.

Us

Principal DS/MLE at small consulting firm.Designed and built C# and .NET toolkit to enable data transfer, aggregation, and analysis between Unity, a Magic Leap AR headset, and AWS infrastructure; this integration supported over 3 successful product launches.Initiated development of active traffic travel time calculations with adaptive signaling; led the client's team of traffic engineers and computer vision experts to enhance urban mobile efficiency.Developed and implemented AWS ETL pipeline to safely process (extracting usable insights using ML models), compress, and transport 20 years of client data into cold storage. This ETL was written using a combination of C#/.NET microservices and python.

Us

Authored in-depth threat assessment of rapid developing technology in transportation, identifying critical vulnerabilities in future controlled hazardous goods transport domestically and internationally.Built a containerized dashboard so clients could analyze traffic saturation flow and seasonality in Washington State; this included a Markov Chain Monte Carlo forecasting model to provide percent-level uncertainties.Lead analyst on employee retention model for federal client. Organized and directed analysis group meetings, and implementation of machine-learning algorithms to increase long term retention of employees using lifelines and h2o gradient-boosted estimators.

Easton, Ma, Us

Professor: Quantitative Analysis, Python, and started the DataLab.

Cambridge, Ma, Us

Post-sales field engineer. Customer success.

San Francisco, Ca, Us

• Worked with Livermore Valley Wine Makers Association and four independent wineries to determineanalysis priorities and potential deliverables.• Created WineFlask, a web app that predicts California wine qualities and score from region and varietal.• Built and validated a multidimensional linear regression model to predict tasting score and utilizednatural language processing to predict tasting notes based on a dataset of 50,000 Wine Spectator reviews.• Deployed web app front end using Flask, Bootstrap, and SQLAlchemy to access a MySQL database.

Livermore, Ca, Us

• Headed analysis team of over two dozen scientists across six institutions to complete a Department of Energy Level 2 Milestone with the first percent-level measurements of fission cross sections. • Modeled non-linear effects of neutron beam and actinide target overlap to improve precision by 2x. • Characterized data quality and streamlined data processing by developing analysis tools written in Python, Perl, Bash, and C++, depending on user and previous system architecture.

Newport News, Virginia, Us

• Guided a tour for Congressional Commission to Review Effectiveness of National Energy Labs. • Tour guide for all experimental areas: halls, beam generation, accelerator, and super radio frequency (SRF) cavity construction area. • Supervised two summer interns from basic coding skills to active analysis for Q-weak in 12 weeks. • Presented experimental results by invitation to 4 international conferences and JLab users group.

Williamsburg, Va, Us

• Developed stand-alone low-current "event-mode" analysis procedure for precision rate extraction • Authored (primary) peer-reviewed paper of the first sub-percent (0.86%) Møller polarimeter results. • Devised and executed experimental plans to extract the parity-violating asymmetry on aluminum to achieve 2% precision. • Created analysis tools and statistical models to understand problems using Perl, C++, Fortran, bash, and Mathematica. Integrated these tools for use with Q-weak experiment’s MySQL database.

Hartford, Ct, Us

• Wrote technical manuals on design and failure modes for eight power generation technologies. • Wrote and presented an hour long webinar on fuel cell technology to a broad internal audience. • Presented at HSB's corporate underwriters' conference on four alternative energy technologies.

Doctor Of Philosophy (Ph.D.), Experimental Nuclear Physics

Masters Of Science, Physics

Bachelor'S Degree, Mechanical Engineering