SEAN BALBALE.
Engineer working where software meets hardware.
About Me
Electrical & Computer Engineering and Computer Science at Trinity College, Class of 2027
I'm a senior at Trinity College, drawn to the problems where software meets physical hardware.
I got into engineering because I wanted to understand how things actually work, not just how to use them. That's led me pretty far down the stack. I've written bare-metal Assembly for an 8051 microcontroller, real-time C++ firmware on an Arduino, and Verilog on an FPGA. For my senior capstone, I'm building a pacing node for rowers and cyclists that predicts fatigue entirely on a Nordic nRF52840, with no cloud offload and under 100 ms of latency. I enjoy working within those constraints, since there's no abstraction layer to hide behind.
I've spent the last two summers on the other end of the stack, in enterprise backend software. At Bullhorn, I owned a Spring Boot 2-to-3 migration across multiple production Java services. At Sonos, I led the move of two production services from WebSocket command delivery to an SQS-backed queue, and tracked down a silent delivery failure in pre-production along the way.
On my own time, I built AstrapheAI, an AI coaching platform for endurance athletes. It pulls training data from Strava, WHOOP, Garmin, and intervals.icu, then merges duplicates into a single record for each workout. From there, a training-load engine calculates fitness, fatigue, and readiness, and a coaching agent uses that history to plan what you should do next. It also has an MCP server that lets you connect your data to Claude, and I've open-sourced the whole platform. It all runs on a four-node Proxmox cluster in my homelab, which I set up and maintain myself, with Kubernetes and a CI/CD pipeline handling deployments.
I also research privacy in the systems we're building toward. My paper, "Embodied Harms and Inferred Data: Redefining Privacy in Extended Reality," argues that current privacy law doesn't hold up for immersive environments that passively collect behavioral biometrics. It proposes a "Bodyright" framework that protects autonomy by design rather than by policy. The paper won the Alumni Prize in English Composition at Trinity's 2026 Honors Day and is forthcoming in The International Journal of Law, Ethics, and Technology.
Outside the lab, I row on Trinity's Division III varsity team, which mostly means early mornings on the Connecticut River. I'm also Technology Chair for the Accidentals, Trinity's a cappella group, where I run live sound and handle recording for our performances. When I have free time, I'm usually cooking something that takes far too long, out on my bike, or ski mountaineering.
EXPERIENCE.
Sonos
Cloud Service Software Engineering Intern
Led the migration of two production Java/Spring Boot services on the Cloud Content team from WebSocket command delivery to an SQS-backed fire-and-forget queue, authoring the design plan and Jira epic, implementing feature-flagged per-player fan-out with retry-until-acceptance semantics, and validating both transports against real speakers. Diagnosed a silent delivery failure in a pre-production environment by tracing commands through the SDK and queue infrastructure to an endpoint-routing defect, and drove the fix across both services. Made the service's functional-test suite runnable locally by porting its DynamoDB-stream Lambda into a local test harness for pre-push validation ahead of the shared CI run, and built Datadog dashboards and monitors for the new delivery path.
Trinity College
Teaching Assistant (ENGR 212)
Instructed students in foundational engineering laboratory practices, focusing on systems analysis and hardware implementation. Guided students through hardware experimentation and objective data analysis to reinforce core ECE concepts.
Bullhorn
Software Engineering Intern
Engineered backend architecture across the Shield and Builders teams for enterprise-scale systems. Owned a major framework migration from Spring Boot 2 to 3, handling complex javax-to-jakarta namespace transitions and refactoring the proprietary data access layer. Led a critical security initiative to upgrade core logging dependencies to Log4j2.
Cherrybrook Networks
Founder & IT Consultant
Founded and managed an IT consulting business providing physical network and hardware solutions. Designed and built custom computers, deployed residential Wi-Fi systems, and automated secure cloud backup infrastructures.
East Coast Metrology
Metrology Intern
Conducted high-fidelity calibration of precision measurement equipment, including large-volume scanners, laser trackers, and portable CMMs. Validated manufactured engineering parts against SolidWorks CAD models with micron-level tolerance to ensure strict adherence to physical design specifications.
Titan Advanced Energy Solutions
Engineering Intern
Architected Python software to parse JSON output from ultrasound sensors. Transformed raw sensor data into waveform visualizations to verify heat maps, directly accelerating the validation workflow for detecting physical defects in electric vehicle batteries.
Weston Public Schools
Engineering Teaching Assistant
Assisted in developing course curriculum and facilitating instruction for high school engineering and physics courses. Maintained and calibrated precision workshop equipment, including FDM 3D printers, laser cutters, and power tools.
SELECTED PROJECTS
AstrapheAI Coach
Built and shipped a mobile-first AI athletic coaching platform — mobile client/web frontend, backend, data pipeline, and infrastructure. Ingests physiological data from Strava, WHOOP, Garmin, and intervals.icu, deduplicates it into canonical records, and computes training load (TSS, CTL/ATL/TSB, a composite Readiness score) server-side to drive a Gemma 4 coaching agent with RAG over training history. Self-hosted on a 3-node HA Kubernetes cluster with an 80%+ test-coverage CI gate; later added a standalone MCP server exposing user data to Claude over a spec-compliant OAuth 2.1 flow, and open-sourced the whole platform under Apache-2.0.
View Details →Embodied Harms and Inferred Data: Redefining Privacy in Extended Reality
The emergence of Extended Reality (XR) represents a massive shift in human-computer interaction, transitioning data collection standards from seeking active user input to the passive, continuous collection of "behavioral biometrics." While existing privacy frameworks (such as BIPA and GDPR) focus on identification and secrecy, this paper argues that they are structurally obsolete in an environment defined by inference and manipulation. By deconstructing the "inference gap," the paper reveals how immersive architectures bypass cognitive defenses to extract sensitive psychometric traits, ranging from health status to sexual orientation, without user awareness. Through a comparative analysis of the Apple-Meta dichotomy, the study demonstrates that market-led "privacy-by-design" functions primarily as an anti-competitive moat—a "Walled Garden" paradox in which privacy becomes a luxury good, and domination is centralized rather than eliminated. Furthermore, by applying the legal doctrines of "undue influence" and "defect of consent" to the "Proteus Effect," the article establishes that XR environments possess a unique capacity to vitiate user autonomy through subliminal behavioral modification. Therefore, the paper proposes a pivot from the "notice-and-choice" legal paradigm to a "Bodyright" framework. This rights-based model reclassifies biometric data not as a raw material for extraction but as an inalienable extension of the human sovereign, ensuring that the integration of the metaverse does not necessitate the commodification of human consciousness.
PROFICIENCIES_
Hardware
- Microcontrollers & Embedded: 8051 Microcontroller, Arduino, Intel Cyclone V FPGA.
- Low-Level Languages: Assembly (8051), Verilog, C/C++ (Embedded).
- Circuit Analysis & Tooling: Oscilloscopes, Logic Analyzers, LTspice, Instrumentation Amps.
- Mechanical & CAD: SolidWorks, FDM 3D Printing, Laser Cutters, Precision Metrology Calibration.
Software
- Languages: Java, Python, TypeScript, SQL, MATLAB.
- Enterprise Backend: Spring Boot 3, Jakarta EE, JUnit 5, Mockito.
- Frontend & Full-Stack: React, Next.js, Tailwind CSS.
- Architecture: Object-Oriented Design, Proprietary Data Access Layers, Zero-Knowledge Architectures.
Cloud & DevOps
- Infrastructure & Deployment: AWS, Docker, Linux.
- Version Control & CI/CD: Git, GitHub, managing feature branches and merge requests to improve pipeline efficiency.
- Security & Logging: Log4j2 deployment, legacy framework migration, dependency hardening.
Let's Talk
Whether you're working on something interesting or just want to connect — reach out. I read everything.