Alan O'Meara · Fractional AI Engineering Leadership
I help engineering teams ship AI agents that survive production.
Most teams can get an AI agent working in a demo. Getting it to run reliably in production — predictable, observable, cost-controlled, and trusted under real load — is a different and much harder problem. That gap is where AI initiatives stall.
Why me
9 years at Adobe, ending as Senior Engineering Manager
Infrastructure engineering and operations for Adobe Campaign — the systems behind Adobe's largest enterprise customers, where downtime gets measured in angry executives and "works on my machine" isn't a defence. DC-to-AWS migrations, high-priority escalations, reliability at scale.
Drove Ketchup to production at Adobe
Led the team and technical decisions that shipped Ketchup — a production AI agent trusted to run inside Adobe's highest-severity (P1/S1) critical-outage channels for Adobe Campaign and Adobe Journey Optimizer. It automated incident summarisation and Jira updates while engineers fought live outages. Not a prototype. Not a hackathon demo. Production software people relied on under maximum pressure, in an environment with zero tolerance for it misfiring.
Founder of Devlar Technologies
Building AI Workforce — an autonomous multi-agent platform. Building a product at the frontier keeps me current on exactly what works and breaks in agentic AI today. The same expertise goes directly into client engagements.
How I help
Demo → production
Take an existing AI agent or LLM feature from promising proof-of-concept to production-grade and reliable. Structured assessment, a clear path, hands-on delivery.
Reliability engineering for AI
Observability, error handling, orchestration, cost control — the practices that keep AI systems trustworthy under real load and give your team confidence in what they're shipping.
Fractional AI engineering leadership
Senior direction for your AI roadmap, architecture decisions, and team without the cost or commitment of a full-time hire. Embedded or advisory, depending on what fits.
How I work
Leadership-led, hands-on where it counts. I work through your engineers — diagnosing, setting technical direction, unblocking, and raising the team's bar the way a senior engineering manager does — and I get into the code myself where it unblocks something fast or sets the standard I want the team to hit. You get the senior engineering leader you're missing, not an extra contract pair of hands.
Who it's for
Engineering-led mid-market and scale-up teams with an AI or LLM initiative that has stalled between demo and dependable production. You have the ambition and some code. What's missing is senior leadership who has actually made agentic AI reliable at scale — someone to steer your engineers, not replace them.
I work with a small number of teams at a time. Engagements are direct — you work with me, not a managed team of juniors.
If you're trying to ship AI that holds up when it matters, let's talk.
Book a 20-minute intro call — no pitch, just a conversation about where your AI initiative is and whether I can help.
Also on LinkedIn