Skip to main content
Bret DuBois

Builds and operates real systems · Redwood City, CA

Technical seller
who builds.

Three years at Apple doing high-volume hardware and software triage: two promotions, 300+ perfect satisfaction scores, one intake conversation that became a $150K+ enterprise account. A year in B2B SaaS running technical discovery for an AI platform. Nights and weekends, I build and operate the systems on this page.

I'm after the work where building a system and explaining it are the same job: solutions and sales engineering, technical account management, and the infrastructure, SRE, and R&D roles behind them. Mostly at AI, networking, and cloud companies. Title matters less to me than whether both halves are on the table.

Status
Open to SE, TAM, infra, and SRE roles
Location
SF Bay Area · hybrid or remote
Focus
AI · networking · cloud infra
In progress
CCNA · AWS SAA
Contact
bretdubois1@gmail.com

Selected work

The fastest way to know whether I can do the job is to look at what I've built. Most of it runs as one system on an always-on box under my desk, managed from the Linux CLI and run like production: version-controlled compose, secrets kept out of git, monitored nightly backups. Here it is as a map; every outlined piece has a writeup.

TelegramJARVIS · Python bot · systemddaily brief · container ops · backup alerts · job digest↗operatesDell OptiPlex · Ubuntu 24.04 · Docker Compose↗postgresn8nopen-webuicouchdb · Obsidian LiveSyncfutures-{ingest,dash,hook} ↗room-inference ↗media stackKVM · Cisco Modeling Labs: CCNA lab topologiesrestic · nightly · monitoredNAS · snapshot repoTailscale meshMacBook · iPhone · gaming PC(GPU for training + inference)
Fig. 01The system as deployed, tap any outlined node to open its case study

The case studies, in the order I'd want you to read them:

RoomTag: indoor positioning from Wi-Fi fingerprintsESP32-C6 · FastAPI · scikit-learn / PyTorch · ONNX · SwiftUIA wearable ESP32 tag scans nearby access points and a self-hosted ML server infers which room you're in, with no cameras, beacons, or GPS. Random-forest and neural models, an inference pipeline with EMA smoothing and Bayesian priors, OTA firmware via GitHub Actions, and an iOS app.Automated futures trading stackPython · PostgreSQL · Docker · broker API · StreamlitThree services that ingest real-time trade signals, parse them into structured intents, run them through a risk layer, and place orders through a futures broker API. Runs in paper mode by design. Includes the postmortem of the incident that took it down.JARVIS: a deterministic-first personal assistantPython · systemd · Telegram · Ollama · local LLMsA Telegram bot that operates my infrastructure: daily briefs, container control, backup monitoring, location awareness, and a job-search pipeline. Designed so every daily-use command works even when every LLM is unreachable.The platform underneath: a homelab run like productionUbuntu · Docker Compose · restic · Tailscale · KVMThe always-on box the other projects live on: a dozen composed services, monitored nightly backups with a written retention policy, a mesh VPN, and written operating rules for making changes safely.Benz Collision: a network core in a closetUbiquiti UniFi · PoE · Comcast Business · Verkada · structured cablingA network redesign for a working auto body shop. Site survey, constraint analysis, and topology design that turns an upstairs closet into the core by reusing camera cabling that already ran through it.

Also built

  • Beat Saber automapper: turns a YouTube link into a playable custom map by wrapping an open-source neural mapper, including three upstream bug fixes (audio decode fallback, codec crash, numpy broadcast) to get it working.
  • RelayCRM: a personal-relationship CRM in Next.js and Supabase with Claude-assisted contact notes.
  • This site: static Next.js on GitHub Pages, hand-set type, hand-drawn SVG diagrams, no animation framework. Design decisions in the colophon.

Work

The building is one half. The other half is six years of translating between technical systems and the people who buy them: enterprise referrals at Apple, technical discovery for an AI platform at SpotHopper, in-home consulting at Asurion. Build the systems, then translate them: that combination is the pitch for a solutions role, and for most jobs where both halves matter.

  • SpotHopper · Business Development Rep · Apr 2025 – Apr 2026
    Technical discovery for an AI marketing platform sold to a skeptical, non-technical buyer base. Prioritized discovery quality over call volume; contributed to the team's highest closed-won revenue month with roughly half of qualified conversations converting to booked meetings. Left deliberately to pursue SE/TAM roles: the AE track there hired externally.
  • Asurion · Field Tech Sales Expert · Sep 2024 – Apr 2025
    In-home technical support and device onboarding: walk in cold, assess an unfamiliar environment, fix it, and explain it in plain language. Exceeded sales quota while keeping customer satisfaction high.
  • Apple · Sales Specialist → Technical Specialist → Technical Expert · Oct 2019 – Aug 2022
    High-volume diagnostics across iOS, macOS, and hardware. 300+ perfect NPS surveys. Spotted an enterprise environment during a routine consumer appointment and referred it: $150K+ in verified lifetime revenue. Mentored teammates on troubleshooting and customer communication.
  • Independent network consulting · 2020 – present
    Design, deployment, and support of Ubiquiti UniFi networks for SMB and residential clients. Case study.

Education & certs

  • B.S. Cognitive Science, Design & Human-Computer Interaction: UC San Diego, 2024. Coursework included data structures & algorithms, databases, networking, and operating systems.
  • A.S. Computer & Information Science: College of San Mateo, 2021, completed while working full-time at Apple.
  • CCNA: in progress: self-study plus hands-on labs in Cisco Modeling Labs running on my own hypervisor.
  • AWS Certified Solutions Architect – Associate: in progress.
  • Eagle Scout: Boy Scouts of America, 2018.

Contact

Email is fastest: bretdubois1@gmail.com. Also on LinkedIn and GitHub. A printable résumé is at /resume, and the thinking behind this site's design is in the colophon.