This case study explores my role in upgrading chatbot experiences across diverse markets, during the years 2023 and 2024. While each of the featured clients had unique needs, the core design process remained consistent. I'll walk you through the key stages of this process, showcasing how User Experience (UX) design principles were applied to enhance conversational design for Artificial Intelligence (AI) interactions.
Challenge
With the rapid advancement of Generative AI, many companies sought to upgrade their existing chatbot and virtual assistant systems. However, simply integrating AI tools like ChatGPT wasn't enough. These systems often lacked a user-centric design, resulting in clunky interactions, inefficient task completion, and frustration for employees. The challenge was to bridge the gap between powerful AI capabilities and intuitive, user-friendly experiences.
Solution
I addressed this challenge by implementing a rigorous UX-driven approach to integrating Generative AI into conversational interfaces. The result was a suite of AI-powered conversational assistants that significantly improved employee efficiency, streamlined internal processes, and provided a seamless user experience.
Clients
My Role
Lead UX designer, I managed and guided a cross-functional team of eight.
Timeline
Overwall: 4 weeks. Discovery & Research: 1 week. Design & Iterations: 3 weeks.
Process
•User Stories•User Flows•Wireframes•Style Guide•Prototype•Development
Tools
•Figma•FigJam•Dialogflow•Vertex AI
Methodology
1. Research 2. Ideation 3. Design 4. Handoff
1. RESEARCH
Kickoff
As the UX design lead, I initiated the project with a comprehensive team alignment via video call, ensuring clear communication and a shared understanding of the client's brief, milestones, and deliverables. Coordinating with the IT team, we collaboratively reviewed initial user flow concepts and technical requirements. To maintain transparency and address potential client inquiries, I established a centralized document for team questions, fostering a proactive and well-informed project kickoff.
Clients requirements
The clients asked for virtual assistants solutions that seamlessly (1) integrated their established brand style and voice, (2) provided a secure environment for their product catalog and content, and (3) streamlined internal processes. Essential to the project were guarantees of (4) high speed and performance, along with comprehensive (5) training for their core teams to ensure successful adoption and utilization.
A critical requirement across all clients projects was the (6) integration of Google's Dialogflow and Vertex AI, driven by existing client partnerships and our IT team's specialized expertise. This necessitated my rapid acquisition of technical knowledge to effectively bridge the gap between UX principles and the specific functionalities of these Google tools. Leveraging my background in web design and development, I facilitated clear communication between the design and IT teams, ensuring seamless translation of UX requirements into technical implementations and vice versa.
2. IDEATION
User Flows
Guided by the established user stories, I led the development of user flows through a collaborative design process. Beginning with individual exploration, I then facilitated group sessions with design peers, content and UI designers, and IT developers to iterate and refine the flows. My primary objective was to ensure each user story was addressed with maximum efficiency and intuitiveness, while seamlessly integrating brand requirements and appropriately handling catalog and content within each scenario.
The final user flow designs effectively balanced user experience and technical requirements, clearly delineating points of Generative AI integration and outlining the anticipated outputs of each interaction. For instance, when requesting a document, employees were presented with flexible delivery options, including PDF download, email delivery or sharing with a specific group or department.
User flow - happy path. Presents all main steps of a full conversation with output.
3. DESIGN
Wireframes
Despite these projects' backend focus, I recognized opportunities to enhance the user experience through strategic wireframing, while ensuring alignment with the client's mandatory use of Google tools like Dialogflow and Vertex AI. I recommended specific improvements, particularly in iconography and button positioning, to ensure intuitive navigation and prevent conflicts with existing app features. With the integration of Generative AI, it was crucial to adapt certain actions, allowing users to seamlessly interact with the new functionalities, all within the constraints of the required Google platform integrations.
UI refresh
To modernize the apps with minimal disruption, we implemented a targeted UI upgrade, introducing a new icon set and refined typography. This refresh aimed to inject greater brand personality, particularly through the introduction of brand-representative avatars for the virtual assistants, enhancing user engagement and recognition.
Prototypes
I developed prototypes of the user flows, to facilitate interactive review and collaboration across teams, incorporating all relevant scenarios and potential variations. Utilizing Figma's collaborative features enabled us to efficiently iterate and finalize design decisions in real-time during team meetings, ultimately streamlining the process and saving valuable development time before implementation in Dialogflow.
4. HANDOFF
Handoffs were done in two phases
First, for front-end enhancements, we delivered an updated style guide, integrated into the client's brand guidelines, and a comprehensive chapter detailing the virtual assistant's identity. This included avatars with varied expressions, tailored to different departments and company events, designed to foster a sense of team integration rather than a purely mechanical interaction.
Second, for back-end development, I personally translated the refined user flows into clear, concise maps, accommodating the AI's technical requirements. This approach facilitated efficient development and streamlined issue resolution, ensuring rapid adaptation to client requests and error corrections.
User flow - happy path. Presents comments and instructions from UX, Content and Dev teams.
Reflection
In these projects, I successfully navigated the complexities of integrating Generative AI into conversational interfaces, demonstrating a user-centered approach that bridged the gap between advanced technology and intuitive design. By leading collaborative teams and employing a structured process, I ensured the delivery of AI-powered assistants that significantly enhanced employee efficiency and streamlined internal processes. The integration of Google's Dialogflow and Vertex AI, coupled with a focus on both UX and technical requirements, resulted in solutions that met the specific needs of diverse clients while maintaining a consistent design philosophy.
Leading the design team for these projects revealed the necessity of fostering team cohesion and focus to minimize duplication and maximize productivity. Given the distributed nature of our team, spanning multiple time zones, we relied on time-independent collaboration to ensure continuous project advancement. I came to the realization that open communication and robust dialogue enabled us to synthesize diverse viewpoints into a cohesive and comprehensive final product. Ultimately, this experience underscored the importance of clear communication, adaptability, and a deep understanding of both user and technical considerations in crafting impactful AI-driven experiences.
