AskIBM

Bringing AI to a
global workforce

AskIBM is a generative AI-powered assistant that redefines workflows for over 280,000 IBM employees worldwide. Powered by watsonx and the Granite LLM, AskIBM streamlines access to IBM’s extensive intranet and tools, offering a unified chat interface that drives productivity. From document analysis to IT and HR support, AskIBM provides intelligent, context-aware solutions tailored to user needs.

My Role

Design Lead

Team

IBM CIO

Start Date

Februrary 2024

👋
What this project highlights

This project highlights my expertise in AI Strategy & Implementation, demonstrating my ability to leverage AI to solve real-world challenges. It showcases my skill in creating intelligent, user-centric solutions that transform workflows, scale innovation, and deliver measurable impact across a global organization.

Team Roles & Responsibilities

The development and success of AskIBM were made possible by a highly collaborative, cross-functional team.

Product Owner

Defined the vision and priorities for the platform, ensuring alignment with organizational goals and user needs.

Design Lead (My Role)

Directed the user experience design, ensuring the platform was intuitive, accessible, user-friendly, and aligned with our AI strategy across IBM.

Engineering Manager

Led technical strategy and implementation, overseeing the development of scalable and reliable architecture.

Iteration Manager

Coordinated agile processes, managed sprints, and kept the team aligned on deliverables and timelines.

Front-End Developers

Built responsive and seamless interfaces across web and mobile platforms.

Back-End Developer

Developed robust APIs and backend systems, ensuring smooth integration with existing tools and data sources.

Researcher

Conducted in-depth user research, usability testing, and feedback analysis to guide design and feature development.

Data Scientists

ptimized AI functionalities, leveraging the Granite 3 large language model to deliver intelligent, context-aware solutions.

The Problem

Designing an AI that
works for everyone

The primary challenge was to harness the power of the Granite 3 large language model and transform it into a cohesive, accessible product that empowers users across the organization. This required addressing several key issues:



Educating IBMers

Ensuring IBM employees understand how to leverage AI in order to get the most out of it for their unique day to day tasks.


Fragmented Tools

Hundreds of teams independently building their own watsonx chat experiences and creating further inconsistency.


Scalability & Consistency

Implementing a unified design language to ensure consistency and usability across web, mobile, and API integrations for AI patterns and components.

Lost Hours

Routine tasks and endless information searches drained valuable time, hampering focus on strategic initiatives.

The Solution

A collaborative approach to redefining productivity

By taking an iterative and user-driven process, we ensured that AskIBM evolved into more than just a tool—it became a foundation for smarter, more integrated workflows that empower IBMers globally.

Understanding the Landscape

Conducted in-depth research to uncover user pain points, identify fragmentation across workflows, and prioritize key areas for improvement.

Iterative Prototyping

Designed and tested prototypes to validate AI functionalities, gathering insights from a diverse user base to refine and adapt the platform.

Unified Strategy

Integrated 13 specialized assistants into a cohesive system, ensuring consistent and seamless experiences across platforms..

Contribution First Framework

Created patterns and components that could be leveraged by our core design systems team and use throughout additional IBM experiences.

The watsonx challenge, a catalyst
for insights and innovation

The watsonx challenge was a company-wide initiative inviting employees to create custom workflows using AskIBM’s workflow builder, with over 3,400 submissions tailored to diverse roles and tasks.

From these, 43 workflows were identified as highly productive and released for broader use, enabling targeted solutions across the organization.

This initiative provided a unique opportunity to gather feedback on AskIBM’s performance, user interface, and results, which we leveraged to refine the product and improve the experience.

The Impact

Transforming work
at every level

AskIBM empowers employees to focus on high-value, strategic work by eliminating frustrations with fragmented tools, improving job satisfaction, and fostering seamless collaboration to boost overall team productivity.

10k

Reported hours saved by IBMers using AskIBM - based on surveys and in depth data analysis.

504k

The average number of prompts and monthly queries inputed into AskIBM since July 2024.

13

Number of dedicated assistants currently federated through AskIBM to support different disciplines and user needs.

3.4k +

Number of community built workflows designed by IBM employees to improve daily tasks unique to their workflows.

Impact beyond the numbers

New Standard in AI Design

AskIBM contributed foundational elements to IBM’s universal AI design library, ensuring consistent and intuitive user experiences across all AI tools.

Building a Scalable Framework

The project established a replicable framework for enterprise-wide AI integration, enabling scalable innovation and seamless unification of tools.

Increase Adoption and AI awareness

Through product tours, blog posts, ongoing tutorials, and the watsonx challenge more IBMers were exposed to the real benefits and time savings that AI can provide.

A Client Zero Story for Sellers

AskIBM’s success positioned IBM as Client Zero, showcasing its AI capabilities and empowering sales teams with a compelling real-world example.

Concepts & Vision

Where it's going

AI will continue transforming digital tools, becoming seamlessly integrated into workflows and interfaces until it’s a baseline expectation. At IBM, we’re focused on overcoming technical hurdles while preserving the existing ecosystem, using an iterative approach to layer AI into search, dashboards, news, and more—paving a natural path to this future.

While this project was focused on the chat experience, the following show some examples of how the tool could be weaved into the fabric of our employee intranet.

Final Thoughts

Reflecting on the work

Getting to work with a tool that builds on AI is an incredible opportunity and experience for any designer. This work demands staying at the forefront of emerging AI trends, patterns, technologies, and understanding the profound impact AI is having on products and society.

Through this project, I learned that AI is fascinating, underutilized, still in its infancy, and here to stay. To design effectively for AI, I realized I must also be a super user of it. By immersing myself in different consumer AI products, I was able to better understand its potential and uncover innovative ways to integrate it into products and experiences.

Think Beyond the Interface

Understand AI’s capabilities deeply enough to design for outcomes, not just features.

Experiment and Innovate

Continuously test, learn, and push boundaries to uncover creative applications of AI.

Stay Curious and Adaptable

Embrace AI’s evolving nature and adapt to the new opportunities it brings to design and problem-solving.

Design for Trust and Clarity

Ensure AI interactions are intuitive, transparent, and aligned with user needs and expectations.