Ciarán O'Hara
AI Product & Design Builder — Local AI, Systems Thinking & Digital Products
San Francisco Bay Area, CA · ciaran.ohara@gmail.com · linkedin.com/in/ciaranohara
Infrastructure program leader and independent AI/product builder developing a portfolio of private AI systems, interactive products, and visual research tools. Combines 20+ years of complex technology program delivery with hands-on experimentation in local inference, RAG, AI infrastructure, UI design, product discovery, and rapid prototyping. Focused on turning complex systems into clear user experiences and practical products.
Solas Sovereign Systems
- Developing a private AI infrastructure concept for organizations that need secure, locally controlled AI over sensitive enterprise data.
- Designed the product around local inference, document ingestion, vector search, RAG, and an on-premises / air-gapped deployment model.
- Developing the product and positioning around AI sovereignty, privacy, infrastructure control, and practical enterprise adoption.
- Exploring secure demonstration appliances and local AI environments using Mac Studio-class hardware and open-source AI tooling.
Global Local AI Foundation / GLAM
- Developed a public-interest initiative focused on local AI, privacy, and environmental sustainability, with Power, Privacy, and Policy as core themes.
- Designed and delivered an Earth Day demonstration showing local AI operating independently of the cloud, incorporating live environmental data and a visible energy / carbon narrative.
- Developed the organization, product concepts, messaging, and web presence around responsible local AI adoption.
White Mirror
- Developed a privacy-first AI concept centered on an elegant, frictionless, and emotionally human user experience.
- Explored the product as a local-first AI environment where sensitive personal interaction can remain under user control.
- Defined UX and product principles emphasizing simplicity, trust, privacy, and human-centered interaction rather than model-centric features.
AI Observatory
- Designed an interactive exploration of the AI ecosystem intended to make models, companies, technologies, and relationships understandable through visual information design.
- Explored ways to turn a rapidly changing technical landscape into an intuitive product experience rather than a conventional directory or database.
- Developed the concept as a polished web experience and research / discovery tool.
Celtic Explorer
- Designed a digital museum and research experience for Irish mythology combining an interactive encyclopedia, narrated stories, character portraits, and source material.
- Integrated mythology, Old Irish pronunciation, etymology, citations, manuscript sources, relationships, search, and filtering into a single exploratory interface.
- Designed the experience around discovery, navigation, and storytelling, including narrated playback and relationship-based exploration.
LegalVault
- Developed a concept for a local encrypted AI appliance for small law firms, focused on private document search, eDiscovery, and retrieval-augmented analysis.
- Explored local vector indexing using FAISS / Qdrant and the use of on-premises AI to avoid sending sensitive legal data to cloud services.
- Developed the concept around a practical deployment model for organizations that need AI capability without surrendering control of their data.
ClarityNow
- Developed a browser-extension concept for simplifying and improving the user experience of enterprise ServiceNow workflows.
- Explored the use of an interface layer over existing enterprise software to reduce complexity without requiring changes to the underlying platform.
- Applied product and UI thinking to a familiar enterprise problem: making complex operational systems easier for people to use.
- Built and tested local AI workflows using Ollama, LM Studio, OpenWebUI, and Hugging Face model ecosystems.
- Explored RAG architectures, embeddings, vector databases, document ingestion, and retrieval using LangChain, LlamaIndex, FAISS, and Qdrant.
- Experimented with local speech and language tooling including Whisper and Piper TTS, and with Apple local AI capabilities.
- Indexed and worked with large document collections, including a corpus of approximately 50,000 court records for vector search experimentation.
- Explored model sizes, inference constraints, local hardware, and the practical trade-offs between cloud and edge AI.
- Developed an AI reading / research workflow using large source collections and structured prompting to investigate emerging AI technologies.
- Developing the Five P Method: People, Picture, Plan, Perform, Pivot — a practical model for connecting systems thinking, visual understanding, and project execution.
- Uses system maps, whiteboarding, Common Operating Pictures, and dependency visualization to turn complex technical environments into shared mental models.
- Applies the same systems-thinking approach to infrastructure transformation, AI products, and emerging technology environments.
- Designs interfaces as working products rather than static mockups, using rapid web prototyping to test ideas in the browser.
- Uses visual information architecture, system diagrams, and interactive prototypes to make complex concepts understandable.
- Combines product discovery, systems thinking, UI design, and technical experimentation to move from idea to demonstrable prototype quickly.
Local AI: Ollama, LM Studio, OpenWebUI, Hugging Face
RAG: LangChain, LlamaIndex
Vector Search: FAISS, Qdrant
Speech: Whisper, Piper TTS
Product / Design: UI prototyping, information architecture, visual systems, browser-based applications
Development: GitHub, web technologies, rapid prototyping
20+ years leading infrastructure and technology programs across Apple, Marvell, Meta, Pearson, San Francisco International Airport, Infoblox, Nimble Storage, Symantec, and Yahoo!. Experience spans data-center transformation, cloud migration, enterprise infrastructure, security, governance, and complex cross-functional delivery. Full infrastructure record: Infrastructure Program Manager résumé →