Kevin Lyons Email & Phone Number

Boston, Massachusetts, Us

- Lead team of 7 engineers working across 5+ Node/TypeScript repositories, deployed via GitOps and Flux- Lead design and implementation of internal tool using react-router to process 25+ exclusion files/week containing PII- Lead weekly agile refinement meetings and provide feedback on all technical plans and pull requests- Lead response for all Production incidents, interfacing with Engineering, Product & Analytics stakeholders- Support hiring via 30+ technical and object-oriented interviews and 8+ campus events to date

Boston, Massachusetts, Us

- Co-led development of Jenkins job to create 15+ new Node microservices, w/ Remix & Fastify templates- Led development of reusable components and theme on top of internal library w/ Emotion CSS- Co-led integration of third-party bot protection vendor, tarpitting 70% of login requests at origin

Boston, Massachusetts, Us

- Led webview standardization initiative, bringing web login and registration experiences to 2.7+ million unique monthly active users across 10+ iOS and Android apps- Led migration to trunk-driven development with automatic deployments via Octo and Bamboo, reducing time-to-merge by 90% to <1 hour

Cambridge, Ma, Us

- MEng, Computer Science, with a focus in Machine Learning and Artificial Intelligence- Worked under Peko Hosoi and Christina Chase in the MIT Sports Lab research group- Completed thesis Automated Force-Velocity Profiling of NFL Athletes via High-Frequency Tracking Data (link coming soon)- Served as TA for 2.98 Sports Technology: Engineering & Innovation course in Fall 2020 semester, mentoring team of 4 undergraduate students- Supported various lab initiatives, including MIT Sports Summit for 2 years, where I led website development both years and the entire event in 2019

Cambridge, Ma, Us

- Coordinated coursework for professional-focused data science and analytics course Topics included are: Supervised and unsupervised learning, clustering and classification, regression and prediction, hypothesis testing, deep learning, networks and graphical models, and recommendation systems.- Monitored course forum to respond to student questions and troubleshoot technical issues- Assisted students with understanding machine learning and data analytics fundamentals- Managed student software workflow using Google Colaboratory, Python, and R- Served 3,000+ students over the course of 3.5 years

Boston, Massachusetts, Us

- Built C# microservice to ingest Zendesk tickets into internal user-level tool receiving 20K+ requests/day- Built React frontend & C# backend to visualize AD groups & roles for all 4,500+ employees

Cambridge, Massachusetts, Us

- Built internal MERN application for manual validation of 100K+ textual records using NLP-powered backend with BERT model- Led integration effort with offsite team to unify and migrate 20+ backend MongoDB schemas - Coordinated application feature requests from 6 internal users in 1-month of production testing

San Francisco, California, Us

- Built distributed data pipelines on Spark for batch processing of 300K+ driver trip records- Integrated with machine learning model that achieved > 0.9 AUC in predicting driver behavior & pricing- Updated Cassandra table schemas and Go microservices to fix a bug that affected 50K+ drivers in NYC

Foxborough, Massachusetts, Us

- Worked with internal personnel to support data-driven analysis efforts using pandas and scikit-learn

Amsterdam, Noord-Holland, Nl

- Streamlined data from CSV/XML/DICOM image formats to common HDF5 data structure- Wrote Python HDF5 data loader for I/O of 20K+ lung nodule crops with preprocessing abstractions- Evaluated deep 2D CNNs for multi-label lung nodule classification use case- Co-authored patented reinforcement-learning-based clinician support system in coordination with MGH

Cambridge, Massachusetts, Us

- Assisted Master’s Thesis student seeking improved data-driven city space allocation for architects- Optimized multi-agent Java simulation of 24 hour real-time to 45 second run-time- Generated train and test sets of 10K+ cities for supervised learning with JSON + Python data pipeline- Utilized linear regression and shallow CNN for traffic volume and solar radiation predictions

Boston, Massachusetts, Us

- Optimized all web service calls by 40% via improved XML serialization in C# backend- Wrote and deployed log monitor in C# to connect cloud computing environment with Splunk + SQL- Integrated load-balanced caching and distributed locking mechanism across replicated web instances- Integrated front end enhancements and bug fixes for improved user interface, while keeping MVC intact

Master Of Engineering - Meng, Computer Science

Bachelor Of Science - Bs, Computer Science

Education record