Hi, I'm
Wesley Dalton
I build.
Software engineer and CS & History student at the University of Pennsylvania. I work across full-stack, machine learning, and cloud infrastructure — most recently at Anduril and A-Life.
Connecting to Spotify...
01 / About
The short version
I'm a software engineer who likes turning ambiguous problems into systems people actually rely on. My work spans full-stack product engineering, applied machine learning, and the cloud infrastructure that ties them together.
Right now I'm a SWE intern at Anduril, and before that A-Life. Along the way I've shipped products at early-stage startups, fine-tuned models for real users, and led engineering teams through Penn's AI Business Club.
I study Computer Science & History at Penn — the second half of that is on purpose. When I'm not shipping, I'm usually deep in a history book or hunting down new music.
University of Pennsylvania
Expected May 2027B.A. Computer Science & History · GPA 3.62 / 4.00
Toolkit
Languages
AI / ML
Infrastructure
Web
02 / Experience
Where I've worked
Software Engineering Intern · Anduril Industries
InternshipMay 2026 — Present · Boston, MA
On the Mission Autonomy — Behaviors & Integrations team, owning an end-to-end rewrite of the autonomous behaviors catalog in modern C++ — redefining how asset behaviors are defined, composed, and reused across the autonomy stack, hardened with unit, FTRT, software- and hardware-in-the-loop testing.
Software Engineering Intern · A-Life
InternshipFeb 2026 — May 2026 · Milan, Italy
Built an AI diagnostics-ingestion pipeline (Next.js, Azure OpenAI GPT-5) that auto-parses 12+ medical report types and 500+ biomarkers into a normalized clinical catalog, plus a real-time collaborative report editor (Y.js CRDTs) and clinical scoring engines feeding an LLM visit-briefing generator.
MOVE AI Fellow · Handshake
FellowshipAug 2025 — Nov 2025 · Remote
Evaluated frontier LLM outputs and designed adversarial test suites surfacing critical model failures for leading AI labs.
Software Engineering Intern · TRAK
InternshipMay 2025 — Sep 2025 · New York, NY
Built an end-to-end order-workflow platform and led a GCP cloud migration — processing 1,000+ daily orders across 50+ sites.
Project Lead · AI Business Club @ Penn
LeadershipJan 2025 — Present · Philadelphia, PA
Lead a 10-engineer team building ML data products for partner companies, including a CRM lead-scoring system that won a 48-hour competition.
ML & Automation Engineer · Civic
InternshipSep 2024 — Dec 2024 · Philadelphia, PA
Shipped a GPT fine-tuning + NLP pipeline that automated message triage and routing across 15+ Congressional offices.
Software Engineering Intern · Flora
InternshipJun 2024 — Aug 2024 · New York, NY
Architected an AI image-generation pipeline and secure billing APIs (Stripe, Clerk) handling 10,000+ daily requests.
03 / Projects
Things I've built
KRAKEN
IEEE OCEANS 2026 · Subsea defense
March 2026 — Present
A multi-agent tethered-UUV architecture for real-time subsea threat interdiction in communication-denied waters. Built with a Cornell · MIT · Draper team at the Albacore 250 maritime-autonomy hackathon — a COTS prototype (Pixhawk/ArduPilot) plus a Monte Carlo framework showing coordinated scouts sharpen localization and intercept rates. Accepted as an IEEE paper at OCEANS 2026 Monterey; presenting in September.
Mise
iOS · Reservations on autopilot
December 2025 — Present
A shipped iOS app that lands tables at the world’s hardest-to-book restaurants. A real-time engine watches Resy, OpenTable, SevenRooms, and DoorDash around the clock and books in milliseconds the instant a seat drops — backed by an encrypted credential vault and Stripe billing. Now serving 200+ customers.
Respire
Burnout prevention, 7 days early
February — August 2025
A full-stack wellness app that fuses WHOOP biometrics — HRV, sleep, and strain — into a multi-signal ML model, flagging burnout risk roughly a week before clinical benchmarks. Pulls live data through the WHOOP API, surfaces trends in an interactive Plotly dashboard, and turns each forecast into concrete recovery recommendations.
InstaLite
A social network that scales
February — May 2025
A production-grade social platform with PageRank feed ranking (Spark), facial recognition (TensorFlow), and real-time chat (Kafka) — scaled to 10,000+ concurrent users on AWS at sub-second latency.
04 / Contact
Let's build something.
I'm always happy to talk shop — whether it's an interesting problem, a full-time opportunity for 2027, or just a good conversation. My inbox is open.