Vilva Athiban P B
I'm a Lead AI Engineer at Omio, one of Europe's leading travel platforms, where I drove AI adoption across the organisation and built the shared MCP services that teams — technical and non-technical — now use every day. I single-handedly built the MCP-first backend of Omio.ai. So my work has two sides: I build the AI infrastructure, and I coach the people and teams who adopt it.
As an AI adoption manager and coach, I don't just ship tools and walk away — I help teams change how they actually work: which problems to point AI at, how to fold it into their delivery process, and how to build the habits that make it stick. On the infrastructure side, I architect the shared MCPs, plugins and skills that give everyone something safe and powerful to adopt in the first place.
Before AI took the seat, I spent a decade building web and mobile products — JavaScript, React, GraphQL, TypeScript, Node — most recently leading UI and Core Platform work at Omio. That engineering background is what keeps both the infrastructure I build and the coaching I do grounded in real production constraints.
What I work on
On the build side: LLM orchestration, Retrieval-Augmented Generation, multi-agent workflows, MCP (Model Context Protocol) architecture, shared MCP services, prompt engineering, and applied ML. On the adoption side: coaching product, data and engineering teams at Omio on where AI creates the most value, folding it into their delivery process, and building the habits that make it stick. I collaborate across the org to find that value — then either build the infrastructure for it or coach the team through it.
Work with me
I take on a small number of company engagements, in the same two roles I play at Omio — builder and coach.
As a builder, in one month I can implement an MCP architecture across your company — a swarm of in-house MCPs, plugins and skills that plug AI into your internal tools, so every team can use AI to the fullest. For example, customer service can see exactly what's wrong with a booking through an MCP wired to your logging tools, instead of waiting on the tech team.
As a coach, I help tech organisations improve engineering and delivery efficiency with AI — driving adoption, streamlining the delivery process and changing how teams work so they ship faster without raising rework. See all services.
Teaching & speaking
I've been teaching developers since 2017 across JavaScript, Python, and Agentic AI. My community has grown to 20,000+ across platforms, and I've delivered 50+ talks across 7 countries. I also design and run hands-on workshops for engineers, product managers, and non-technical stakeholders who want to move from AI curiosity to practical application.
Beyond that, I write for LogRocket, Medium, and dev.to. Recent pieces have been about Agentic AI, MCP, and how tools like Claude Code use React under the hood.
Outside of work
I'm originally from India, living in Berlin. I paint when I need a break from screens, travel whenever I can, and spend plenty of time on movies and meals with friends.
The International Flower
One long-running side project: I collect a small amount of soil from every country and region I visit. When I cross the 25-country mark, I'll mix it all together and seed a flower plant in it — so the flower that blooms will have grown from soil gathered around the world. A tiny, slow-motion piece of travel made living.