Full-Stack Agentic AI Engineer
Multi-Agent Systems · Responsible LLM Systems · Operational Decision-Support
I build agentic AI systems that hold up outside a demo, and the full-stack applications around them: data model, secure execution, source-grounded reasoning, evaluation, release. Sole developer on a commercial platform, plus client software and my own multi-agent tooling. Everything below you can open, run, or ask me to walk through.
Cybersecurity, then systems, then design, then data. It reads like a detour until you see what I build now.
My path wasn't a straight line, and the convergence is the point. It started in defensive security (an NSA/NSF-funded cybersecurity camp, then a City of San Diego internship), ran through a Management Information Systems degree with a Computer Science minor at San Diego State, and picked up a designer's eye during a 250-hour internship in Rome. MIS taught me how systems serve a business; CS gave me the depth to build them.
From there it went through data and machine learning, then into full-stack consulting, and now into building agentic AI systems for commercial use. Each step kept the one before it. The security instinct decides how a system fails. The MIS training decides what the system is for. The design year decides whether anyone can use it. That combination is the specialization, not four separate resumes stapled together.
The work keeps a consistent shape whatever the domain: applications with real data models, testing and CI; AI that stays grounded in sources it can cite and hands consequential decisions to a person; and analysis that says out loud when a headline number is misleading. The claims stay checkable. Live links and public repos where they exist, honest labels where the work is private, and nothing rounded up in either case.
Software should earn trust, and that is an engineering property, not a promise. It comes from what the system is prevented from doing.
I design agentic and multi-agent systems around deterministic control instead of open-ended autonomy. Work advances in dependency-ordered waves, each one isolated and gated behind automated tests, a build-and-boot check, and repository-state checks, so a failure is contained where it happened rather than propagating into later state. A human still holds anything consequential.
An answer is worth what its sources are worth. I build reasoning layers that carry provenance and uncertainty all the way to the interface rather than flattening them into confidence, verify claims against the record instead of letting a model fill the gap, and treat model output as untrusted input behind an explicit allowlist. Then I evaluate it, because a demo is not evidence.
Schema to interface to release, with security built in from the start rather than patched on: strict Content-Security-Policy and HTTP security headers, no secrets in the client bundle, least privilege by default, per-user isolation enforced in the data layer. Shipped through automated pipelines running lint, tests, build, and secret scanning, on pinned, lockfile-enforced dependencies.
Four builds that carry the most weight: a commercial platform, my own multi-agent tooling, a full product, and a live app. Each one opens to what it does, what it is built on, and how it was verified.
A commercial agentic decision-support platform I built as sole developer, from canonical data model to interface.
Mission ControlMulti-agent AI development orchestration: parallel coding agents, isolated execution, dependency-ordered waves behind automated gates.
META nutrition and training platform built end to end for a working strength coach's practice.
Client systems, research, and analytics. Smaller in scope than the four above, and labeled exactly as far as each one actually got.
Validated Research LibraryA governed-AI knowledge system turning unstructured expertise into a source-verified, queryable library.
Governed Operations SystemAn AI-assisted operational decision system for a small-business client, with a human committing every change.
HydrOSA reinforcement-learning simulation of a multi-stage water-filtration system, validated before hardware.
mediCalm
CDC Diabetes-Risk Analytics POCAn academic ML proof-of-concept classifying three-class diabetes risk on a large CDC public-health dataset.
AlignFlowA multi-source data-quality / master-data-management engine for reconciling records into one trusted view.
Also built
Full-Stack Secure Agentic AI Engineer
Sole developer on a commercial agentic decision-intelligence platform, from the data model up to the interface. I owned the Python services and domain logic, the React and TypeScript front end, the APIs between them, and the canonical schema everything resolved against, while the requirements were still moving. The reasoning layer had to be source-grounded: every value it surfaced kept its provenance and its uncertainty instead of being flattened into a confident answer, and a human approved anything consequential. I built the evaluation to match, because on this kind of system you cannot ship on a demo. Release ran against 22 independently graded acceptance families, a 2,807-check graded suite, and 151 end-to-end browser assertions. The client and their data stay confidential.
Independent Consultant
My throughline is trustworthy AI. Since 2024 I have run an independent practice turning ambiguous business goals into software people actually run their operation on, which mostly means deciding what the system is allowed to be sure about. The anchor is a fitness-coaching platform I own end to end: a Next.js and TypeScript build over a 49-model Postgres schema, with decision engines for progression, scheduling, and nutrition, held to 319 automated tests, secure per-client isolation, and a human coach reviewing every consequential call. Around it sit a source-grounded content system for a physical-therapy practice that published 100+ deliverables behind a hard citation gate, and an AI-assisted operations system for a small business that turned prospects into recurring weekly work and cut a core workflow by about a third. I also build my own AI-native development tooling, which is where the multi-agent orchestration work lives. Clients get fast adaptation without giving up the parts that make a system safe to depend on.
Commercial Operations, Data Analytics & AI Solutions Intern
At Celltrion I built the analytics and decision-support layer for a commercial-operations team inside a regulated pharmaceutical environment. The centerpiece was a production master-data report in Power BI, split into an operations view and a sales-validation view, that teams leaned on every week across all 50 states and the U.S. territories. I also stood up anomaly detection over ten-thousand-plus records with a triage workflow that sorted issues by frequency and impact, and I scoped an Azure OpenAI integration with the secret-handling and human-review gating a regulated setting demands. It was where I learned what production really means when a real business depends on the numbers being right.
UX/UI Design Intern
A 250-hour design internship in Rome, where I designed and built three client websites in WordPress and Figma on an international team. Working across languages and stakeholders taught me the part most engineers skip: how a system communicates its purpose is part of the engineering, not a coat of paint. I still design with that in mind.
Cyber Security Intern
My first technical role, a cybersecurity internship where I worked through Linux and defensive-security fundamentals and helped run the learning operations for the intern cohort. It is where a habit started that never left: think about how a system fails before you think about how it ships. That instinct now shows up in every threat model and test suite I write.
Education
San Diego State University
Graduated May 2026 · San Diego, CAB.S. Management Information Systems
Double minor in Computer Science and Interdisciplinary Studies. Weber Honors College, cum laude (GPA 3.52). Coursework included graduate-level Reinforcement Learning (CS 654) and Machine Learning (CS 549), data structures, computer architecture, linear algebra, and statistics.
University of San Diego
May 2019 · San Diego, CAGenCyber Cybersecurity Summer Program
An NSA and NSF-funded academic cybersecurity camp: defensive security primitives, threat modeling, and secure system administration. This is where the security habit started.
I take on two kinds of work, and I am open to both right now.
Full-stack AI, agentic AI, AI product engineering, secure AI systems, and high-ownership forward-deployed work. The kind of role where someone has to own a problem end to end, talk to the people who have it, and ship something that holds up in front of them.
For businesses and founders who need a real software system rather than an AI demo: manual processes, disconnected spreadsheets, or an idea that has to become something you can run an operation on. I handle the full build.
San Diego County or fully remote. Tell me the role or the problem and what you are building.
