Available for remote work
Applied AI, workflow systems, and technical implementation
I turn messy real-world workflows into working software: multimodal pipelines, automation systems, developer tooling, and AI products with real state, review, guardrails, and delivery.
Selected Work
Original systems built from scratch with real workflow logic, real constraints, and real operational behavior.
A staged backend workflow that turns radiology PDFs into patient-facing video artifacts, with physician review required before rendering. Built as a practical multimodal system with extraction, classification, script generation, review, rendering, and delivery rather than a prototype that stops at text output.
A backend automation system structured around sessions, commands, and normalized event delivery. Built with concurrency controls, throttling, replay behavior, and realtime output streams so it could support sustained workflows instead of just looking good in a demo.
A structural memory layer for coding agents that parses code with Tree-sitter, stores dependency graphs in Neo4j, and adds semantic retrieval on top. Built so an agent can reason about codebase structure and blast radius instead of falling back to keyword search.
A product for streamers where viewers spend Twitch Bits to trigger TTS messages in a cloned voice. Built around live event ingestion, payment gating, voice generation, and realtime audio output with creator monetization as the product goal, not just voice synthesis for its own sake.
How I work
This is the part of the job I am best at: understanding how work actually happens, then building the system that should exist.
I start by mapping data flow, state transitions, and failure modes before writing implementation. Components, APIs, storage, queues, and review gates need to be placed intentionally. Code is easier to revise than structure.
A lot of useful engineering lives at the seams: connecting external APIs, LLM providers, media services, databases, and queues into something that behaves reliably end-to-end. That is where many systems actually break, and where I am most comfortable.
I care about why a system should be built a particular way, not just how. That means reasoning about scope, maintainability, and operational reality. A technically clever solution that is brittle or overfitted to a demo is still the wrong solution.
I optimize for getting something into a real, runnable state quickly. Rough edges can be refined once the system exists and can be observed. Systems that never ship cannot be improved. That is not a defense of slop; it is a view about where effort should go first.
Background
I did not come up through a CS program or a standard bootcamp path. My background is in healthcare, medical training, research, teaching, and customer-facing operations.
I moved into software by building directly: real systems in Python and TypeScript, hands-on architecture decisions, and projects that had to run end-to-end instead of staying theoretical.
I do not treat the nontraditional background as a liability to apologize for. Healthcare and research training built product judgment: understanding workflows, reasoning about real-world constraints, and recognizing when a technically clever solution creates more problems than it solves.
I am especially effective in healthcare-adjacent AI and workflow products, where domain understanding matters alongside the engineering and where "move fast" carries real consequences.
Role fit
Remote-first roles where ambiguous requirements, applied AI, workflow design, and technical judgment all matter at once.
Building products that integrate LLMs, multimodal models, and AI services into working end-to-end workflows. Not research theater. Actual product.
Designing and building automation pipelines: event ingestion, processing, normalization, delivery, and operational controls.
End-to-end ownership of feature development with product judgment — not just implementation tickets handed over a wall.
Customer-facing or internal implementation work: integrations, custom workflows, system configuration, and technical onboarding.
Pre- or post-sales technical work where product understanding, workflow design, and clear communication matter as much as engineering skill.
Building tools for developers or technical users — agent interfaces, workflow automation, CLI and API-first tooling, internal platforms.
Contact
Available for remote roles. If you need someone who can get close to the workflow, understand the constraints, and build the actual system instead of just talking about it, I would like to hear what you are working on.