IBM Consulting Advantage

Making enterprise AI easy to trust and use across IBM Consulting

Role

Senior Product Designer

Timeline

November 2024 - Present

Scale

160,000+ Consultants, globally

01 / Overview

Advantage gives IBM consultants powerful AI tools… but the experience was fragmented and opaque, so consultants didn’t trust or know how to use them.

02 / My Role

I redesigned how consultants see, trust, and move between AI capabilities, from reasoning transparency to a unified chat and library experience.

03 / Challenges & Proposed Solutions

Challenges

Opaque outputs
Disjointed chat flows
No visibility for team leads

Proposed Solutions

Reasoning stream that shows the model’s thought process and sources
One multi-model chat where switching and discovery are simple
Created team analytics dashboard

04 / Key decision

Show the thinking, not just the answer

I ran 25+ hours of interviews with my Lead UX Researcher across seniority levels and found a clear pattern: consultants didn’t trust how AI outputs fit into their work. So I prioritized transparency over automation, designing a reasoning stream that shows how the model got to its answer, including its steps and sources.

Figure 1: One of the features we rolled out was our reasoning and thinking abilities during deep research. Users are now able to see a more concise stream of consciousness and the thought process that the model is able to take while answering.

Figure 2: created enablement gifs to help enable our users to understand how to navigate through

some of the new features we rolled out.

05 / Results & impact

The reasoning stream, unified multi-model chat, and team analytics dashboard I designed are now live across IBM Consulting Advantage, the platform used by 160,000+ consultants.

Trust in AI outputs: Consultants can now see the steps and sources behind an answer, instead of getting a result they have to take on faith.

One chat instead of many: Switching models and discovering tools now happens in one place.

Visibility for team leads: Team owners can monitor activity, model and agent usage, and performance from one dashboard, so they can make adoption decisions with data.

“Your approach to problem-solving is incredibly impressive. Regardless of the design challenge, you consistently present multiple robust and well-considered solutions that spark valuable discussions amongst us. It's clear that your thought process involves thorough analysis and consideration.”
Susan Aboela-DaSilva, Senior Product Design & Research Lead, IBM Consulting

The reasoning stream, unified multi-model chat, and team analytics dashboard I designed are now live across IBM Consulting Advantage, the platform used by 160,000+ consultants.

Trust in AI outputs: Consultants can now see the steps and sources behind an answer, instead of getting a result they have to take on faith.

One chat instead of many: Switching models and discovering tools now happens in one place.

Visibility for team leads: Team owners can monitor activity, model and agent usage, and performance from one dashboard, so they can make adoption decisions with data.

Susan Aboela-DaSilva, Senior Product Design & Research Lead, IBM Consulting

06 / Reflection & next steps

1. Start with consistency when a product has had no designer in 5+ years.
Before any new feature could land, I had to align patterns, the Carbon design system, and four teams working with four PMs. That groundwork is what made the new features feel like one product instead of a collection of parts.

2. Design for the newest user and the power user at once.
Consultants ranged from AI beginners to heavy users, so I paired transparency features with onboarding and guidance instead of building for just one group.

What's Next

I’m continuing to refine the reasoning and chat experiences and helping shape the next generation of the platform. The goal is the same one that drove this project: making AI accessible to IBMers and clients worldwide in a safe, ethical, and impactful way.