Recently at Ozmo, I fostered psychological safety to navigate an AI-driven team evolution and identified early user-need driven opportunities, leading to Ozmo's first AI implementation, a Conversational Self Support Assistant that answered 400% more tech support questions.

Learning through safety

As AI crept into the software industry as a predominant, disruptive technology - I knew my team and I needed to adapt and evolve. I found and allocated time they could dedicate to learning and experimenting with this new technology and tools. I personally researched and found emerging design communities, learning resources and courses to facilitate our needed education.

In this AI era, I believe Psychological Safety is paramount for Design Teams. I facilitated this by increasing the amount my team and I communicated around our learnings and set early experimental goals so expectations existed. I led routine discussions where we shared our successes and failures. I identified those that were champions of the new technology - whether it was prototyping or continued advocacy for human-centered approaches.

I also ensured those who had concerns about the emerging space could be heard and wove their concerns into our approach concertedly. I guided brainstorming and led visioning on how AI could be strategically leveraged across our products or in totally new solutions in user-centered ways.

Here's an example of how I quickly leveraged our new AI skillsets to implement Ozmo's first AI-powered feature:

💼 Case:

Ozmo's Self-support product previously consisted of a stepped-disambiguation workflow to find tech support answers, meaning users had to express their technology first before their issue.

⚠️ The problem:

From a UX perspective, users generally approach problem solving with their issue as their starting point.

🛠️ Process:

I led deep research into understanding effective Conversational Design as I see it as the root of modern interactivity with AI technology. I leveraged effective communication techniques like Active listening, paraphrasing, dedicated pacing, and more as the backbone of an internally defined practice and approach to any AI implementation at Ozmo going forward.

I identified this known user-need (expressing their issue first) as an early implementation opportunity for introducing AI technology into our products.

With this research and new practice defined, I led my team to a new version of our Self-Support product where users could now begin their troubleshooting journey by expressing their issue first.

View a detailed look at the Conversational Support Assistant work ⇨

✅ Results

This new AI experience led to a 400% increase in the product’s delivery of answers to users.

The transformation didn’t end there - it continues as does Design adapting to the new technology and changing product landscape. Next, I prioritized a new wave of work with our Design System to prepare it for AI enablement. While the system itself was completely modernized with tokenization and collaboratively adopted by Engineering, it was clear that the Design System itself would become a layer of governance as more and more cross-functional colleagues were building AI prototypes. This new wave of expression and experimentation needed to be governed, guided and supported by Design, not hindered.