[Case 03]
Turning Enterprise Knowledge into an On-Demand AI Agent
Consulting
Thought Leadership Knowledge Assistant
Accelerating Research and Insight Discovery with AI
[Project Overview]
Led the end-to-end design of TLKA (Thought Leadership Knowledge Assistant), an AI-powered research tool that streamlines how KPMG professionals discover and leverage the firm’s thought leadership. Partnered with cross-functional teams to deliver an intuitive, efficient user experience, enabling rapid access to insights, client stories, services, and alliances that enhance client engagements and strategic decision-making.
[Problem Statement]
KPMG professionals often face time-consuming and inefficient research processes when preparing for client engagements, requiring them to search across multiple systems and sources to find relevant content. The existing methods can slow down productivity and limit the ability to leverage KPMG’s comprehensive thought leadership effectively. TLKA addresses this challenge by consolidating content discovery into a single AI-driven platform, enabling quicker access to accurate and up-to-date information, ultimately enhancing client preparation and internal efficiency.
[Industry]
Consulting
[My Role]
Lead Designer
[Platforms]
Desktop and Mobile
[Timeline]
November 2024- January 2025
[Persona]
John Roberts
Senior Manager, Advisory Practice
I just want to find the right insights quickly so I can focus on preparing for my client meetings.
Age: 38
Location: Chicago
Tech Proficiency: Moderate
Gender: Male
[Goals]
Quickly find relevant KPMG thought leadership and case studies for client presentations.
Trust the platform to deliver accurate, up-to-date information without having to cross-check multiple sources.
Save time by streamlining the research and preparation process for meetings and proposals.
[Frustrations]
Wasting time searching multiple systems and websites for relevant insights.
Uncertainty about whether the information is current and reliable.
Inconsistent formats and lack of clarity in available content.
[Process]
[01] User Research
Conducted interviews with 12 advisory professionals to understand their research and preparation workflows.
Analyzed user behavior in existing research tools to identify bottlenecks and pain points.
Held feedback sessions with client-facing teams to understand their content discovery challenges before important client engagements.
[02] Insights
Users were frustrated by the amount of time it took to locate relevant, current insights.
Professionals needed a single, trustworthy source that consolidated multiple content types (insights, services, case studies).
Users wanted simple, intuitive search functionality that returned precise results without excessive filtering or refinement.
[03 Design Solution]
Designed a unified, AI-powered search experience that delivers quick, accurate access to KPMG’s thought leadership.
Created a streamlined user interface with suggested prompts and follow-up questions to refine searches.
Integrated source files and links into responses to boost user confidence in the credibility and currency of content.
[04] Testing & Iteration
Conducted usability testing with 30 advisory professionals to validate ease of use and relevance of search results.
Ran A/B testing comparing the new TLKA interface against legacy search methods.
Collected iterative feedback through WalkMe and in-tool surveys, refining the design based on real-time user input.
[Outcomes]
Reduction in time spent searching for relevant content during client preparation.
Increase in user satisfaction with research tools, as measured by post-launch surveys.
Higher engagement, with 60% of client-facing professionals adopting TLKA within the first three months of launch.
[Key Learnings]