Gabriel Arnold
Software Engineer
I build full-stack systems and applied ML tools — from distributed monitoring infrastructure and geospatial intelligence platforms to encrypted protocols and real-time web applications.
About
I'm a software engineer and recent CS graduate from Dickinson College with experience shipping production systems across three engineering internships.
Most recently at AVEVA, I built a platform-wide synthetic monitoring system in k6, TypeScript, and Kubernetes to cut service-degradation detection time across 29+ microservices and enabled 200+ engineers to self-serve monitoring through JSON configs and Grafana dashboards — avoiding $12,000+ in annual tooling costs per environment. Before that, roles at WebstaurantStore and Lambgoat grounded me in RESTful API design, database optimization, and CI/CD workflows.
Outside of work I build things I find genuinely interesting: a geospatial ML platform scoring data center siting viability across the US, an end-to-end encrypted messaging system that transmits over sound, and an open-source campus shuttle router with an LLM-powered assistant.
Experience
Software Engineer Intern
- Reduced detection time for service degradations across 29+ microservices and cut on-call escalation load by building a platform-wide synthetic monitoring system in k6, TypeScript, and Kubernetes.
- Avoided $12,000+ in annual costs per environment by leading a tradeoff analysis across 5+ monitoring solutions and designing a system around an open-source stack (k6, Prometheus, Grafana) instead of Azure's native tooling.
- Enabled 200+ engineers to self-serve monitoring through JSON configs, OpenTelemetry, and Grafana dashboards, architected in an 11-page design document developed with 4 cross-functional teams.
- Secured synthetic traffic across all monitored services by implementing secret management in Azure Key Vault, with token caching and automatic refresh to meet reliability and security compliance.
Software Engineer Intern
- Engineered 4+ RESTful API endpoints establishing a single source of truth for warehouse address management, improving data consistency across enterprise applications.
- Improved operational visibility for 900+ warehouse packers by integrating barcode scan metrics into 3 key packing interfaces.
- Cut stored procedure calls by 98% by optimizing critical shipping and carrier data retrieval with caching strategies.
Software Engineer Intern
- Grew monthly pageviews from 750K+ to 1M+ by re-architecting the frontend and backend for mobile responsiveness using ASP.NET Core 3.1, SQL Server, Bootstrap, and jQuery.
- Generated $10K+ in first-year sales by leading development of an e-commerce store with secure PayPal checkout integration.
- Deployed a 97% accurate ML.NET model to automatically detect and remove malicious user content, improving platform security.
- Accelerated delivery time by 67% by establishing a CI/CD pipeline for automated testing and production deployments.
Projects
Geospatial ML platform scoring 214K+ H3 cells across the US for data center siting viability. 25-step feature pipeline over 18 public datasets; LightGBM classifier at 0.971 AUC-ROC with top-100 precision 72× better than random; interactive deck.gl frontend with configurable scoring weights, live in production.
Open-source, campus-agnostic shuttle routing system configurable for any institution. Computes multi-leg walk-and-shuttle itineraries via the Haversine formula, renders on Google Maps, and exposes an MCP server powering an LLM-based shuttle assistant over SSE.
End-to-end encrypted acoustic messaging CLI — no network required. X25519 key exchange, AES-256-GCM authenticated encryption, commit-reveal handshake preventing key substitution attacks, and a Short Authentication String for MITM detection. 64-test pytest suite.
Extensible RAG framework combining Pinecone vector search with PostgreSQL conversation history for context-aware chatbots. Configurable memory and similarity thresholds reduce redundant context retrieval.
Civic app for anonymously reporting reckless driving with photo evidence. AI license plate recognition auto-extracts plate numbers; WebSocket-powered dashboard delivers real-time report updates to law enforcement.
Full-stack public records search platform for property and court docket data. .NET 8 Clean Architecture with CQRS and MediatR; React/Redux/Vite frontend with Google reCAPTCHA and automated record ingestion.
University library management system with role-based auth via ASP.NET Core Identity, Entity Framework Core on SQL Server, and a responsive Bootstrap admin interface with sortable paginated tables.
Vaccination tracker visualizing US COVID-19 progress. .NET Core backend aggregating CDC data every 30 min, MongoDB storage, GitHub Actions deploys to AWS Lightsail, and a Twitter bot publishing automated real-time vaccination graphs.
Education
Dickinson College · Carlisle, PA
Object Oriented Design · Algorithms and Data Structures · Analysis of Algorithms · Software Engineering · Open Source Software Development · Database Systems · Computer Networks · Operating Systems · Computability and Complexity · Mobile App Development · Data Mining · Discrete Mathematics
Conrad Weiser High School · Robesonia, PA
Computer Science 1 · Computer Science 2 · Advanced Topics in Computer Science · ACSL Concepts and Programming · AP Computer Science A
Departmental Award — Computer Science (top graduating senior)
Awards
Awarded to the project that best embodies what the Dickinson faculty would love to see students accomplish. Recognized for ShuttleKit.
Contact