📍 Camas, WA / 📧 [email protected] / 🛠️ dusktreader@github
Passionate Engineering Manager dedicated to delivering quality software while leading my team with velocity and empathy. Seeking growth opportunities for personal and professional development within vibrant collaborative working environments.
- Languages: Python, TypeScript/JavaScript, Perl, C++
- Technologies: FastAPI, Flask, Angular, Docker, Kafka, SQLAlchemy, Spark, Postgres
- Platforms: ClickUp, JIRA, Slack, Google Workspace
Omnivector Solutions: Engineering Manager
- Developed complete Engineering Development process including comprehensive workflow diagrams and extensive documentation of the entire engineering lifecycle.
- Oversaw migration from Monday.com to ClickUp and established workspaces and workflows adapted to the Engineering Development process.
- Established SCRUM-based methodology for engineering team to coordinate development, measure progress, and continuously improve engineering process.
- Instituted regular One-on-One meetings with 8 direct reports to track professional development and establish constructive candor in team communication.
- Conducted performance reviews for direct reports to assess progress and establish improvement goals.
- Directed development of the Vantage HPC SAAS platform from ideation through public-facing deployment in production environments.
- Produced specs and project plan for enabling subscriptions to the Vantage platform through the AWS Marketplace.
- Refactored Jobbergate OSS project to integrate with Vantage platform including development of server-side agent with remote Slurm job submission capabilities.
- Developed custom git branching model for Vantage projects based on the Stable Mainline model to support multiple concurrent deployments of the product and maintain repository consistency.
- Lead effort to produce technical documentation for the Vantage product by contributing copy for many sections and making quality improvements throughout.
Office Ally: Web Development Team Lead
- Piloted effort to integrate lab reports and processing statuses from genomic testing venture project into Patient Ally.
- Collaborated with Architecture team to develop a Kafka-based event driven architecture for moving genomics related data and status changes throughout cross-platform ecosystem.
- Revised Sprint Retrospective meeting structure to improve team engagement, encourage reflection on team dynamics, and promote positive feedback among team-members.
- Handled challenging transition to Team Lead in the midst of dramatic turnover, performance challenges, and team re-direction.
- Codified and regularized One-on-One meetings with direct reports to track professional development and establish constructive candor in team communication.
- Rallied an exhausted and discouraged team around a last minute crunch-time project mandated by high-level stakeholders and delivered results while maximizing engagement.
2019 - 2020: Patient Ally Frontend Rewrite
- Lead effort to rewrite legacy frontend application from Django to Angular 11 using modern best practices and reactive user experience
- Introduced ngrx store for system-wide application state management and reactive asynchronous communication with backend subsystems.
- Re-designed PA Messaging section from scratch to behave like familar email applications including contacts, attachments, and intuitive conversation-based message presentation.
- Promoted version control workflow standardization and collaborated with other senior developers to codify standards and best practices for git.
- Mentored junior devs struggling with new technologies and reactive design to improve overall team velocity in conversion project.
Comscore: Senior Software Engineer
- Spearheaded and lead development of mission critical internal system for record linkage and entity resolution (deduplicating and disambiguating metadata) from disparate subsystems of Comscore.
- Developed flexible graph-like data schema in Postgres to support heterogeneous data models in a single large metadata store.
- Architected ETL pipeline to consume and normalize multi-format data from a wide range of sources including S3, FTP, and external databases.
- Designed scalable recommendation engine using Spark for distributed processing to deliver aggregated link recommendations utilizing multiple, independent matching algorithms.
- Forged stand-alone API application using Python and Flask to service front-end UI as well as direct access by external services.
2012 - 2016: Refactor legacy record linkage systems
- Launched efforts to consolidate several mechanisms of “title matching” (human curated record linkage driven by algorithmic heuristics) for the On Demand Essentials product.
- Designed modular, data-driven subsystem to centralize record linkage and facilitate rapid deployment of new heuristic algorithms.
- Extended logging and event bookkeeping to support forensic analysis of failure events.
- Ported new MCP subsystem from On Demand Essentials to Digital Download Essentials creating a reusable code base that could be extended to similar products.
- Orchestrated crisis management effort for the new subsystem when a major data-provider began delivering invalid data, and used new capabilities to triage problems.
Batelle Inc.: Applied Parallel Computing Scientist II
- Anchored modernization team for open-source biomolecular software project funded by National Institute of Health.
- Translated extensive legacy FORTRAN 77 computing backend into equivalent-performance ANSI C with custom minimal object oriented abstraction layer.
- Commuted entire project from Subversion version control system to Git to improve management and re-integration of exploratory source-code branches.
- Converted build system from Autotools to Cmake to improve stability and flexibility of build system while reducing overhead and improving maintainability of build scripts.
- Collaborated with premier university research team specializing in computer vision.
- Converged with two off-team developers to transition exploratory research code written in MatLab to deployment ready platform implemented in C++.
- Designed distributed computing platform for computational tasks including a dispatching system for client requests from a custom Qt client-side application to a compute cluster using ZeroMQ for low-latency inter-process communication.
- Researched and developed object detectors that hunted for ~2500 pixel objects in ~10 gigapixel images utilizing OpenMP to locally optimize searches.
- Experimented with signature-based classifiers for disjoint object categories using Boosted Tree Classifiers, Support Vector Machines, and Artificial Neural Networks.
- Explored signature generation in C++ with OpenCV using morphology, curvature metrics, and structural statistical moments.
- Developed a novel hybridization of a Self-Organizing Map with an Artificial Neural Network to classify images by structural signature of sub-features.
2004 - 2008: Washington State University: Computer Science B.S.
- Graduated Cum Laude
- Focused on Scientific Computing, Machine Learning, and Parallel Computing