Moving to an AI enablement team
Moving to an AI enablement team
On June 16, 2026, I moved to an internal AI enablement team. It felt like a good milestone, so I wanted to write down what changed and what I hope to work on next.
Before the move
I had mainly worked on large-scale data platforms:
- Building pipelines that process billions of records every year
- Developing a platform that makes management information available across the company
- Reverse-engineering and rebuilding a legacy system that had become a black box
I was often seen as “the data person,” but I had always wanted to become an engineer who can write solid backend systems.
What comes next
At the AI enablement team, I plan to focus on the following areas.
Evaluating GraphRAG
I am evaluating whether GraphRAG is actually necessary for a real-world use case. The important part is not assuming GraphRAG is the answer, but comparing it with a standard RAG baseline across quality, cost, and latency. The platform is built on Google Cloud.
Backend for daily profit reporting
I am designing and building the backend that makes daily operating profit visible in a dashboard. I am involved from requirements gathering through API design, OpenAPI contracts, response shaping, authentication and authorization, and implementation.
It has been especially rewarding to work across the whole path from design to code. The stack includes Hono, OpenAPI (zod), Drizzle, and vitest.
How I want to work
When the move was decided, I told my manager clearly: I want to work on backend engineering.
That does not mean I want to stay away from AI. I think combining “data × AI × backend” could become a strength that is uniquely mine. Rather than stopping at being “the data person,” I want to move toward being a backend engineer who can design systems and write the code.
This move gives me a chance to build the AI part of that combination through real work. I want to take it one step at a time and turn the experience into durable skills.
Closing
My role is still taking shape, and the current phase is fluid. The new environment is a little intimidating, but I am genuinely excited about having more ways to contribute. I will write another update when there is more to share.