The Challenge
Create a proof of concept for executives that showcases how Accenture can utilize the power of AI to enhance workflow
Timeframe: 1 month
Team: 1 UX designer, 1 visual designer, 1 design director, 1 Associate MD
Role: Design Director
Our approach
Identify how to make a meaningful experience with real users
Our team rapidly conducted three focus groups to start to define the vision for our proof of concept and brainstorm use cases with select employees that have recently worked with AI.
Participants came from a variety of backgrounds and career levels to understand a broad range of tasks and experience.
Personality & Unintended Consequences Workshops
To understand how employees would like to communicate with AI and any potential risks, we got the team together to workshop ideas for a range of personality components including tone, power dynamics, level of personalization and character traits. We also assessed various risks to both the user and the business to ensure we are soliving in advance for key issues.
Before designing screens, the team prioritized key uses cases to demonstrate value and drafted a storyline that reflects a potential user journey. Then we conducted landscape research to collect samples from comparable experiences in market to gather inspiration and inform interaction patterns.
Utilization of Accentures newest design system helped the team stay on brand and move faster in developing fully designed screens.
Initial Output
Landing page allows for quick access to ask a question, provides suggested prompts based on previous behavior and highlights recommended exploratory content informed by the users role and history
The user engages with an AI agent that answers questions using internal and external sources. The agent can ingest new documents on request and update responses accordingly, with sources adding credibility and enabling deeper exploration.
The user requests to generate a story based on the response and an agentic huddle occurs.
The agentic huddle represents a group of AI agents that are carrying out various tasks to generate the story on behalf of the user. This is shown as a quick loading screens while the story loads.
Once huddle is complete, the story is displayed for the user. They are then able to view sources in more detail, and start co-editing the content with the agent.
In addition to editing content via hover, the agent can support the user in identifying key subject matter experts, top document contributors or industry leads to reach out to and get their feedback.
Mobile capabilities including voice conversations with AI allow for quick feedback. Our agent listens to the SMEs input and provides suggested updates. Some can be automatically incorporated or shared with the original author for validation.