CX Cloud AI driven insights
Enhancing the delivery of critical insights to improve the user experience in Cisco CX Cloud
I am leveraging AI to deliver network advisories in a more engaging and conversational tone, aiming to drive better insight consumption and adoption. The goal is to communicate more efficiently with minimal friction. Additionally, I added AI-generated video "stories" as an alternative, more compelling way to present these insights, making the information more accessible and impactful for users.
ROLE
Concept, UX, UI Design
DATE
July 2024
Overview
When we open our laptops and connect to the internet, we don’t usually stop to think about how it all works—unless something goes wrong and we find ourselves offline. In a corporate environment, it’s the network engineers who ensure that the network runs smoothly, securely, and efficiently. Their job comes with a lot of responsibility, and they often juggle multiple tasks at once.
Managing thousands of devices in a large corporation is a huge challenge. Engineers need to maintain inventories, manage warranties, track security vulnerabilities, and ensure that devices are up to date. That’s where Cisco CX Cloud comes in—a cloud-based SaaS portal designed to help customers optimize their Cisco equipment. CX Cloud automates inventories, tracks devices, and surfaces critical insights, ensuring users can get the most value out of their Cisco investments.
Problem statement
While CX Cloud provides valuable insights, users expressed frustration over how difficult they were to find. The insights were buried in a separate tab and displayed in filterable tables. The experience felt dry and utilitarian. This resulted in missed opportunities for action and slower response times.
How might we surface insights so users can get them immediately but also have the flexibility to ignore them when focused on other tasks?
Process
My task for this project was to reimagine how AI-driven insights could be delivered in a way that’s intuitive, engaging, and easy to act on.
Although CX Cloud already provided the raw data in a format engineers were familiar with, there was a significant opportunity to enhance clarity and reduce cognitive load. Important updates were often buried in static tables, making it easy for users to overlook essential information. Leveraging Generative AI, I saw a chance to transform this data into concise, context-rich summaries that were easier to understand and act upon.
The broader question became: Could we create a system that not only improved communication but also established a feedback loop between Cisco, customers, and partners, driving better outcomes across the board?
I began by sketching out multiple concepts to explore how we could present AI-driven insights in a more conversational, approachable way. Early on, I focused on:
Conversational Summaries – How could AI transform technical details into updates that felt more like casual check-ins from a colleague, while still being actionable?
Prioritization & Engagement – How could we ensure that users only saw what mattered most, but could dig deeper when they wanted to?
I progressed creating sketches to low-fi to mid-fidelity prototypes, validating and iterating them internally with engineers, peers and stakeholders.
How insights are currently displayed
Initial sketch
Solution
After gathering initial ideas, I created low-fidelity wireframes that focused on displaying insights in a concise, headline format. Similar to how a colleague may present the information in a stand-up meeting. For instance, instead of listing six devices needing firmware updates, we could present it as: “There are six new firmware updates affecting 56 routers, most located in the west coast warehouses.”
This format allows users to understand the issue at a glance and decide whether they need to dive deeper. I added links for users who wanted to explore details further. This allowed us to strike a balance between keeping insights visible without cluttering the interface.
Having worked at Cisco for years, I recognized the value of a tool that could communicate network status quickly and clearly. By surfacing these insights on the landing page and using natural language, we could save users time and help them make faster decisions.
Current version
How insights are currently displayed
1st iteration
Initial sketch
I proposed implementing this in two phases to gather feedback and refine our approach. First, we would surface the insights directly on the dashboard in the new conversational format. This phase would allow us to validate the concept, tone, and content with users.
In phase two, we plan to offer an alternative delivery method using AI-generated video avatars. This would provide a more dynamic, immersive experience for users who prefer video over text, allowing them to watch insights delivered as short "stories" narrated by AI avatars.
Below are some of the mock-ups I created in Figma to illustrate this concept. These mock-ups showcase how insights will be displayed prominently on the landing page with links to deeper details.
2nd iteration
Integrate AI-generated avatars and expand options to address each insight directly from the homepage.
Insights delivered as AI generated video
Present information in natural, conversational language, as if explaining insights and their value to a colleague in the room.
Users can have a video call with the AI agent via Webex.
Or interact via chat.
Mobile version
“This look great and way above what Cisco normally offers in such a space.”
Outcomes
Once implemented, a way to measure success is to track if this feature improves retention rates by helping users act on insights faster and more effectively.
Surfacing useful insights whilst making them easy to understand and follow up can reduce churn, improve retention and potentially help with customer acquisition. Concept and planning were extremely well received when presented, highly praised by the creative approach and alignment with business strategy. This aligns with Cisco’s broader vision of leveraging AI to create a more immersive, engaging user experience.
The concept was well received by peers and management, who praised its alignment with business goals and its innovative approach. We are now moving forward with plans to explore AI video avatars and prepare for further user testing.