Gemma Hall

UX + AI CONTENT | MULTIMEDIA SERVICES

CAPGEMINI

For an independent thesis project in partnership with Capgemini UK and their innovation and design consultancy, frog, a cutting-edge conversational sales trainer called EdenAI was developed. This solution was designed to support the professional development of entry to mid-level staff within the organisation, providing a unique, AI-powered platform for sales training. At its core, EdenAI utilises two advanced Large Language Model (LLM) agents (Adam and Eve) trained on comprehensive sales theory. These agents enable users to engage in simulated real-world sales conversations, replicating authentic scenarios that help sharpen their skills and improve their confidence. By providing dynamic, real-time responses, EdenAI creates an interactive and adaptive learning environment that is both practical and engaging, tailored to the needs of sales professionals. The platform was designed using a variety of low-code tools, including Soul Machines for creating conversational AI experiences, integrated with OpenAI’s ChatGPT 3.5 for enhanced natural language processing, and built on the Wix platform for ease of access and scalability.

The development of EdenAI was grounded in a robust and multi-faceted research methodology to ensure the platform addressed the specific needs of Capgemini’s staff. This process involved user surveys and in-depth interviews, gathering insights from the target audience, including entry and mid-level sales staff and their managers. These conversations revealed key challenges and areas for development, informing the platform's functionality and features. Ethnographic studies were

also conducted to gain a deeper understanding of Capgemini’s organisational culture and the daily realities faced by its sales teams. This allowed the design of

EdenAI to align with the company’s existing workflows and sales training approaches.

The overwhelmingly positive feedback during the pilot phase has led Capgemini UK to move forward with the implementation of EdenAI into their broader sales training programmes. This step marks a significant evolution in how sales training is delivered, showcasing the transformative potential of AI-driven solutions in professional development.

8/10 participants expressed that the model showed great potential as a valuable training tool.

7/10 reported that the prototype was more interactive and engaging than pre-existing sales training platforms at Capgemini.

8/10 participants indicated strong interest in using the AI agents for their ongoing professional development.

10/10 supported the complete development of a future iteration for application within Capgemini.

90% of participants acknowledged that AI could not entirely replace human coaching, however emphasised that it would serve as an effective supplementary tool to enhance traditional sales pitch training.

RESEARCH FINDINGS

View Eden AI Prototype

EdenAI is not only an innovative solution for sales training but also a reflection of the growing potential for AI to transform corporate learning. By leveraging low-code platforms such as Soul Machines, OpenAI’s ChatGPT 3.5, and Wix, EdenAI demonstrates how AI-driven technologies can be integrated with ease and efficiency into existing corporate structures. This project highlights the importance of creating adaptable and scalable tools that can evolve alongside the needs of an organisation. As Capgemini moves forward with the implementation of EdenAI, the platform is set to play a critical role in improving sales

performance, boosting staff confidence, and reshaping the way professional training is delivered. The success of this project opens up future possibilities for AI applications not just in sales, but across other learning and development areas within large organisations, setting a new standard for interactive, AI-assisted

training.

UX PROCESSES

User Research & Persona Development:

Conducted user interviews and surveys to understand the training needs of entry to mid-level staff, focusing on the challenges they face in upskilling and adapting to new technologies.
Developed personas representing key users, such as junior professionals looking to improve technical skills and mid-level employees seeking to enhance efficiency through automation.

User Journey Mapping:

Mapped out the journey of users from discovering the platform to completing AI-powered training sessions and applying their skills in the workplace.
Identified user needs, motivations, and potential pain points, such as ease of use, personalised learning paths, and the effectiveness of AI-driven agents in their training.

Wireframes & Prototyping:

Designed low-fidelity wireframes in Figma to structure the user interface, focusing on intuitive navigation, training module access, and integration with the AI agents.
Created high-fidelity interactive prototypes to demonstrate key functionalities, such as automated task training, real-time feedback, and progress tracking for skill development.
Trained the large language model (LLM) powering the platform using datasets tailored to entry to mid-level job roles, ensuring the AI agents are capable of providing relevant and accurate guidance.

Usability Testing:

Conducted usability testing to assess the platform's ease of use, the effectiveness of AI interactions, and the quality of the learning experience.
Gathered feedback from participants, who highlighted the benefits of real-time, personalised guidance from the AI agents and the platform's ability to help them improve their skills efficiently.