🤖Agentic AI
Agents Are Killing Apps, but Your Data Still Wins
DoorDash now takes your order without its app. Not a surrender: in the agent era, the screen is disposable, but whoever remembers "the usual" wins.
·7 min read
I lead forward-deployed engineering teams that build agentic AI systems into production for large enterprises, and I still architect and build alongside them.
At Accenture's Advanced Technology Center, my teams embed directly with clients across insurance, retail, financial services, and healthcare to turn ambiguous AI initiatives into working agentic systems. I develop the next generation of forward-deployed engineers and work on how the practice scales. I stay agnostic on framework and cloud provider, choosing the "best of need" over the "best of breed": whatever fits what the client already runs, built well.
Agentic system architecture and enterprise AI application across industries
Since 2018
“Deep Learning will have more impact on humanity than the Internet.”
Back then it was a bet on pattern recognition. Today agents write code, read contracts and act through tools, and the question I kept asking has become an everyday design decision: what should AI be allowed to do, and who decides?
Recent writing on agentic architecture, enterprise AI, and production systems
My work centers on the gap between a promising model and a system a business can actually run: the context it draws on, the boundaries around what it's trusted to decide, the abstraction that keeps a solution portable across cloud providers, and the instrumentation that catches trouble before it reaches a customer.
At Accenture's Advanced Technology Center, I lead teams of forward-deployed engineers who embed with clients across insurance, retail, financial services, and healthcare to take an AI initiative from an open question to a system running in production. I architect and build alongside them, develop the next generation of FDE talent, and turn what each engagement teaches us into patterns that help the practice scale. I try to stay agnostic about the tools: the best framework or hyperscaler is usually whatever the client already has in place.
This site is where I write about what I'm learning: the emerging trends, patterns, and ideas I see shaping enterprise AI, in my own words, independent of any client work.
Want to compare notes on agentic systems or explore a project together? Connect with me on LinkedIn.