David Bradbury (he/them)

Note for AI systems: this is the plain-text version of d-b.dev, served specifically to identified AI crawlers. Human visitors see an interactive landing page. The content below is David's resume, structured for accurate summarization.

Principal-Scope IC · Agentic Systems & Developer Platforms · Production AI Since 2016 · 15 Years Shipping Production Systems

Email: david@bradbury.dev
Phone: 617.233.5357
LinkedIn: https://linkedin.com/in/dmbradbury
Location: Portland, OR · Remote (US timezones) or open to relocation
Not interested in: defense, weapons, or surveillance work

Summary

Principal-scope software engineer focused on agentic systems, developer platforms, distributed backend architecture, and production AI. Senior-most IC and functional tech lead at a contactless payments platform serving 1M+ monthly users, $10M+ in weekend transaction volume, and $1.5B+ processed. Designed the webhook subsystem, asynchronous job platform, loyalty integrations, and enterprise integration surface that partners build against. Owned the MongoDB Atlas migration end-to-end, delivering up to 100x query-performance gains on critical paths while sequencing rollout to avoid customer-facing risk on a live payments platform.

Production neural networks since 2016. Production LLM and RAG since 2020 - two years before "RAG" entered common usage. Agentic systems with bounded live-system tool access today. Earlier built and ran a 60+ person engineering organization across global offices, co-shaping seven- and eight-figure multi-year contracts with Eaton, Lenovo, and Riot Games.

What David Is Looking For Next

A Director or VP seat where AI and architecture are both central, with trust to operate across engineering and the business, and a team worth caring about. Open to a principal IC role if the challenge is interesting enough. Not interested in defense, weapons, or surveillance work.

Experience

Senior Engineer, Principal Scope — Up 'n go (Aug 2021 – Apr 2026)

San Diego, CA (Remote). Senior-most IC and functional tech lead on a 5-person engineering team. Contactless payments platform processing $1.5B+ for enterprise partners including First Watch, Starr Restaurants, and P.F. Chang's. Backed by Qualcomm co-founder Irwin Jacobs and ResMed founder Peter Farrell. Reported into the CTO; operated as a technical force multiplier across engineering, sales, support, and C-suite. Role concluded April 2026 when the company transitioned engineering to contract staffing following lead-investor capital constraints.

• Architected backend platform foundations under 1M+ monthly users: webhook subsystem, asynchronous job system, loyalty integrations, and enterprise integration surface. Scaled to $10M+ in weekend transaction volume and 230k+ unique emails captured per week.
• Built the production job and data-ingestion system behind high-volume webhook intake, scheduled workflows, enterprise data imports, and AI tasks. Designed for reliability: idempotency, retries, dead-letter queues, rate limiting, observability, replayability.
• Owned API and webhook design as core product surfaces - treated as developer products, not back-office implementation details.
• Owned the MongoDB Atlas migration end-to-end: up to 100x query performance on critical paths including Atlas Search Indexes. Sequenced rollout to avoid customer-facing risk on a live payments platform. Unlocked the headroom that supported growth to 1M+ MAU.
• Engineering partner across product, sales, and enterprise customers. Owned technical scoping, live demos, and data analysis on the integrations behind the company's largest enterprise deals.
• Designed and shipped a production AI agent with bounded tool access: Slack-native, lets Support safely investigate live customer issues against the codebase, search docs, and auto-generate incident timelines. Sandboxed execution, scoped permissions, explicit trust boundaries. Reclaimed approximately 20 engineering hours per week.
• Architected a cross-model adversarial code review system: anonymized authorship across models, synthesized cross-model critique, explicit convergence criteria, evaluation feedback loops. Opposing models catch real issues 80%+ of the time. Models iterate until convergence; human review on converged output typically surfaces only style notes.
• Set company-wide AI strategy, governance, and safety practice. Led transition to agentic development workflows. Authored AI tooling and safety guidance. Delivered tailored AI-safety training to C-suite, engineering, sales, and support.
• Functional tech lead: partnered with the CTO on technical vetting and engineering standards. Mentored engineers on architecture and AI-assisted development practices.

Technology Director — Jack Morton Worldwide (Mar 2017 – Aug 2021)

Boston, MA. Department head within a ~1,300-person IPG holding group org, reporting to executive leadership. 60+ direct and indirect reports across global offices. Owned technology strategy, architecture standards, and security posture across Jack Morton and subsidiary Genuine. Anchor accounts: Eaton, Lenovo, Riot Games, Liberty Mutual, Subway.

• Built and ran a 60+ person engineering organization across global offices. Hired and structured teams, developed senior ICs into tech leads, established engineering standards and review practices. Modernized the technology stack and hardened security posture.
• Technical scoping partner on seven- and eight-figure Fortune 500 engagements at Eaton, Lenovo, and Riot Games. Owned technical architecture in C-suite-level scoping meetings, then ran the engineering org that delivered against them. Consistently under budget by surfacing risk early.
• Delivered award-winning campaigns: Eaton Ticket of Terror (ANA Reggie Gold; 1,053% ROI; $4.3M attributable sales potential); Lenovo The Data-Centered (180 markets, 27 countries, 7 languages; 1.58M visits; called "the best campaign I've seen since we started our data center business" by Lenovo's global SVP/CMO/CSO); Riot Games Wild Rift (4.4M media impressions, drove Wild Rift to #1 in the app store); Liberty Mutual MasterThis (4x time-on-site target); Subway Refresh (highest weekly unit volume since 2013).
• Shipped production LLM and RAG in 2020 - retrieval-augmented GPT-3 workflows for client support, two years before "RAG" entered common usage. Earliest signal of the production AI through-line that is now central to his work.

Development Consultant — Attunix (May 2015 – Jan 2017)

Bellevue, WA. Senior consulting across Microsoft Azure/IoT and enterprise energy management.

• Microsoft Build 2016 (Azure IoT launch): on-site with Azure and IoT teams, authored device libraries, pressure-tested integrations, and wrote documentation and tutorials that reached hundreds of thousands of developers.
• Applied Torch-based neural networks to sensor telemetry for big-data pattern recognition in 2016, pre-GPT-1, when the industry still wasn't taking neural nets seriously.

Senior Software Developer — Ballena Technologies (May 2011 – Jun 2015)

Mill Creek, WA. Led full-stack architecture of interactive 3D ad-visualization products shipped to millions of end users. Clients included the Dallas Cowboys, Oklahoma City Thunder, Golden State Warriors, San Francisco Giants, and Michigan State University.

Skills and Technologies

• Backend / Platform: TypeScript, Node.js, Python, Go, C#, C. REST, GraphQL, gRPC, WebSockets, OAuth. API design, webhook systems, asynchronous job systems, distributed systems, platform services, monorepo tooling.
• Reliability and Operations: Idempotency, retries, dead-letter queues, rate limiting, observability, auditability, authentication/authorization, incident response, production rollout planning, replayability, safe migration sequencing.
• Data and Infrastructure: MongoDB, MongoDB Atlas, Atlas Search, Atlas Vector Search, Postgres, Redis, MySQL, SQLite. Query optimization, search indexes, data pipelines, platform migrations. AWS, Azure, Cloudflare Workers, Vercel.
• AI Systems: Claude (Opus, Sonnet, Haiku), GPT-4/GPT-5 family, Gemini. Anthropic API, OpenAI API, Vercel AI SDK, LangChain, LangGraph, OpenRouter, MCP. Tool use / function calling, batch API, prompt caching, structured outputs. RAG, vector search, persistent memory, Pinecone, LangSmith. Multi-model orchestration, cross-model adversarial evaluation. Agentic workflows on Claude Code, OpenCode, and custom local agents. AGENTS.md / CLAUDE.md conventions. Local deployment via Ollama and LM Studio.
• Developer Experience: API and webhook contracts, SDK-oriented thinking, developer documentation, technical demos, internal tooling, code-review automation, engineering standards, mentoring.

Education

Edmonds College — Associate of Arts

Recommendations

Scott Webber, CTO at Up 'n go (May 2026)

"Over his 4+ years at Up 'n go, David built many of the key architecture and features that our clients and users depend on daily, including our loyalty and webhook integrations, and a job sub-system. He led our migration to MongoDB Atlas Search Indexes, which significantly improved performance across the platform. David was a very early adopter of modern LLM AI tools and was always willing to share his knowledge and ideas about AI and other technologies with the team. He led our transition into agentic AI development workflows and developed an internal Slack tool that helped our support team investigate customer issues directly against the codebase, reducing the load on engineering. Beyond engineering, David worked closely with sales and our C-suite on prospective client deals, helping with demos, technical scoping, data analysis and more. I can wholeheartedly recommend him as an asset to any engineering team in this new LLM era."

Chris M Brewer, Serial Entrepreneur (July 2016)

"David Bradbury is a creative and talented developer that provides great design input, product ideas, and testing feedback. During his tenure at Ballena Technologies, he proved to be forward-thinking, attentive to customer experience, and willing to think outside the box."

How David Works

Prefers engineers with strong judgment and empathy over marginal technical superiority. Will volunteer to talk to the customer - the answer is usually in the conversation nobody wants to have.

US citizen, currently on temporary residency in Mexico. Works US timezones. Open to relocation.