IBM

The problem

Client Engineering exists to co-create with IBM’s global client portfolio, which means the design problem is rarely the interface. It’s getting a room of people who don’t share a vocabulary — a client, data scientists, engineers, sales — to agree on what’s actually being built, fast enough that the building still has time to happen.

I was a senior designer intern on the Global Sales, Client Engineering team, and both of my main projects had that shape: a hard deadline, a mixed team, and no shared starting definition.

What I did

The reusable part was the process, not any one screen. Discovery to frame what we were even looking at, framing to narrow it to a decision, solutioning to commit — run as workshops rather than as documents, because a workshop produces agreement and a document produces comments.

IBM Harvest came out of the internship hackathon: 72 hours to an MVP tackling global waste. I covered user research, the app interface, visual design, and branding — which in a three-day build means deciding quickly what not to design.

The capstone was a chatbot for a client, built with data science and engineering interns: a taco catering business that needed to answer customer questions and convert enquiries into orders. I led the design thinking workshops, did front-end development, and prototyped.

Outcome

Two MVPs shipped inside four months, both with mixed-discipline teams and both against a fixed deadline. The methodology diagrams above are the transferable part — they’re the reason a 72-hour build produced something coherent rather than something merely finished.