01 / Client view
Graphic Designer Portfolio
A compact, client-ready graphic designer portfolio prototype for web designers.
Open project viewConsultantAI Builder
Seung ParkBuilding the Future.
Full-stack engineer, now working at the AI-engineering end of the stack. Leading architecture operations by day; shipping one AI-powered SaaS platform by night, with coding agents doing much of the typing and a harness that keeps them honest.
- Currently
- Senior Consultant at Capco, serving as Architecture Operations Lead for a Large Electric Utility.
- Independently
- Piloting SpandaNada, a multi-tenant studio-management SaaS with RAG agents on AWS Bedrock. Around it: a cross-project SDLC harness that governs how coding agents work (hooks, gates, telemetry) and a local-model agent fleet on Apple Silicon.
- Active missions
- 04in flight
- Time on station
- 18.4years
- Crew
- Solowith agents · open to ethical AI work
Identity · 18+ years in the field
About me.
Full-stack engineer with a growing focus on AI application development and 18+ years in technology, spanning enterprise architecture, program management, and consulting. My mindset is Kaizen: always learning, always improving, and working toward more challenging problems.
Currently leading operations for an Application Architecture team at a U.S. electric utility while independently building two SaaS platforms: a multi-tenant studio management platform with AI-powered RAG agents and a teacher productivity tool with constitutional AI safety guardrails, both built with a fleet of autonomous coding agents running under my own SDLC harness.
In 2025 this was experimentation; it's since become production discipline, shipping multi-tenant architecture, RAG pipelines, and constitutional AI safety layers, and running a self-governing agent fleet to build them. Focus stays the same: AI-powered tools that actually help people.
Ethical AI, Sans Doom
How do we use AI to create a better future for all: sustainably, fairly, and with real utility?
Building AI that helps people while respecting our planet and humanity: useful, accountable systems over hype or vibes 🙄
Thinking back at my Contemporary Issues in Computer Science coursework circa 2005, the topics we covered (automation, ethics, AI, and societal impact) feel more relevant than ever. Every AI company needs to establish their own constitutional framework. Transparency in AI values isn't optional; it's essential for building systems we can trust.
My goal is to build brilliant AI that actually helps people, without torching the planet or our humanity.
- 01People-first utility
- 02Planet-aware compute
- 03Accountable by design
Mission log · 04 active
Bodies in orbit.
Prototype gallery · client viewing
Client projects.
Comet trail · 18 years
The path here.
Skills snapshot
What's on board.
AI engineering on top of eighteen years of full-stack and enterprise architecture work: agent systems, multi-tenant SaaS, cloud infrastructure, and the operating discipline to run them.
Core competencies
- AI engineering & agent systems
- Multi-tenant SaaS architecture
- Full-stack TypeScript
- Cloud infrastructure & DevOps
- Architecture operations & program leadership
Recent practice
- Agentic coding workflows · harness, hooks, gates
- RAG agents in product (Bedrock + pgvector)
- Constitutional AI safety layers
- Local model serving on Apple Silicon
- Autonomous multi-agent fleet orchestration (Herdr)
- Spec-first delivery with negative-control gates
Working stack
- TypeScript · Node.js · Python · Bash
- Next.js 16 · React 19 · Tailwind
- Fastify · Drizzle · Aurora PostgreSQL · pgvector · Valkey · BullMQ
- AWS Bedrock · ECS · Lambda · DynamoDB
- OpenTofu · GitLab CI/CD · Docker
- Claude Code · MLX · LM Studio · Herdr
- Stripe Connect
- Vitest · Playwright · Turborepo
- UiPath RPA
Open channel · downlink active