Applied AI Engineering & Systems Design for Creative Technology
We turn creative AI ideas into production systems. ONE Systemics designs and builds AI workflows, tools, and infrastructure for entertainment, music, immersive media, and spatial computing. We work where AI models meet creative software — where reliability, latency, provenance, human control, and real-world workflows matter. We design, build, and hand off systems your team can actually operate.
What We Build
AI Workflow Engineering
Human-in-the-loop AI systems for creative and technical workflows.
- AI-assisted creative workflows
- Agentic workflows and MCP infrastructure
- API and tool orchestration
- Custom plugins and integrations
- Local-first AI tools
AI Systems & Infrastructure
Architecture connecting models, data, APIs, agents, creative applications, and production systems.
- AI system architecture
- Model and API integration
- Local and hybrid AI
- Data and provenance pipelines
- Real-time and offline processing
- Evaluation, safety, and testing
Prototype → Production
Focused engagements that take an idea from technical exploration to something your team can actually use.
- Prototype sprints
- Fixed-scope integrations
- Custom AI tools
- DSP / ML systems
- Production deployment and handoff
System Audits
Understand what’s working, what’s getting in the way, and what to build next.
- Architecture review
- Workflow analysis
- Performance and latency
- Model and infrastructure evaluation
- Licensing and dependency review
- Implementation roadmap
Selected Work
- GroovyUI: A patchbay for local AI audio workflows.
Visual workflow tooling for connecting local AI audio models and creative processing components.
View on GitHub → - MCP2OSC: MCP infrastructure for creative software control.
An MCP server connecting AI systems to OSC-enabled creative tools.
View on GitHub → - LLM4OSC: Profile-bound AI control with safety metrics.
Constrained LLM-to-OSC control with safety mechanisms and golden tests.
View on GitHub →
How We Work
Understand → Design → Build → Validate → Hand Off
We start with the actual workflow rather than an abstract AI use case.
We choose the simplest architecture that solves the problem, build against real requirements, validate with real workflows, and document the result so your team can operate it.
Built to Hand Off
We don’t want to become another layer your team has to manage. Our goal is a system that works without us.
Engagements
- System Audit — Architecture and workflow review with a prioritized implementation roadmap.
- Prototype Sprint — A focused build to determine whether an AI workflow or technical concept actually works.
- Integration Sprint — Connect an existing model, prototype, or technology to your real production workflow.
- Advisory — Technical guidance on AI architecture, models, infrastructure, and creative technology.
All engagements are fixed-scope where practical. No open-ended hourly billing.
Let’s Build Something Useful
Have a workflow you think AI could improve? Bring the workflow, a sample asset, an existing system, or even a half-formed idea. We’ll spend 30 minutes understanding the problem and determine whether an audit, prototype, integration, or advisory engagement makes sense.
Email: onesystemics at gmail dot com
Substack | Youtube | Discord
Location: Los Angeles
Founder and principal engineer: Yuan-Yi Fan, PhD