IN DEVELOPMENT

Transforming "Tara"

From chatbot to dynamic cybersecurity copilot

About

"Tara" is a browser extension designed to prevent cybersecurity risks in real-time by intervening when employees are about to engage in potentially risky actions.
During Savvy's first days, Tara was designed as a simple chatbot to test and prove the concept. However, the initial chatbot concept faced significant limitations, so it was time to take it to the next level.

Company

Savvy Security

Role

Product Design Lead

Time frame

2023-2024

Problem statement

The chatbot was often misunderstood by users, who mistook it for part of the website they were browsing.
Furthermore, the static nature of a chatbot didn't align with our diverse use cases.
Additionally, the old chatbot did not differentiate visually between dangerous and non-dangerous situations.

Solution

The redesigned Tara is a dynamic, adaptive cybersecurity copilot. The modal can take on various forms depending on the situation: from simple informational pop-ups to interactive walkthroughs or chat interfaces. The detached status bar ensures that users are always aware of Tara's presence and purpose, with easy access to settings and information.

Tara’s "moods" are a color system visually indicate the level of risk or action Tara and the user are in. This system allows to provide real-time, context-sensitive feedback. Tara’s mood can change dynamically within a single interaction, providing instant visual feedback as users navigate potential threats.

Tara’s visual design is friendly, approachable, and helpful, avoiding a “big brother” vibe. The design system supports Tara’s dynamic character, with consistent logic applied across color palettes for all moods, ensuring a cohesive and reassuring user experience.

Tara's different faces

Learning

A/B Testing: Old vs New
The dynamic modal led to an improvement in user understanding and acceptance: Users appreciated the flexibility and contextual relevance of Tara’s new design. The new design also led to a significant reduction in attempts to bypass Tara, and more users chose to follow Tara’s recommendations, demonstrating increased trust and engagement.

The positive response was a clear indicator that the redesign successfully addressed the limitations of the original chatbot approach.

Key takeaways
The evolution from the original RidePlan to RidePlan 2.0 successfully addressed the growing complexity of Via’s services.

However, the introduction of advanced features also highlighted the need for comprehensive onboarding and continuous user support. As RidePlan continues to evolve, ensuring that users can fully utilize its capabilities will be crucial to maintaining high levels of satisfaction and operational efficiency.

Want to hear more about this project?

Let's get in touch.

lukalaraf@gmail.com

Let's get in touch.

lukalaraf@gmail.com

© 2024

Made with ♥

Let's get in touch.

lukalaraf@gmail.com

I'm Luka

Product Designer based in Tel Aviv.

I love turning complex tasks into simple designs.

This portfolio shines best on desktop 🌟
Hop over for the full experience.

Need more details?