This week on the GeekWire Podcast, my guest is Brad Anderson — a former Microsoft corporate vice president who spent more than 17 years at the Redmond company. For nearly five years, he has been with Qualtrics, where he’s president of products, user experience, engineering, and security.
I’ve known Brad for a while. In fact I’ve been a guest on two of his shows back when he was a leader in enterprise mobility and cloud technology at Microsoft, when it was all the rage for executives to have their own video shows.
There was “Lunch Break with Brad Anderson,” where he drove his guests around the streets of Redmond in his Tesla. And there was “The Ship Room,” where he talked about the cloud transformation and played games with guests, challenging me to distinguish real startups from fake ones.
Qualtrics, the experience management technology company, has dual headquarters in Seattle and Provo, Utah. Once publicly traded, it was acquired in a private equity deal in 2023.
For this week’s show, I went to the Qualtrics Tower in Seattle to talk with Brad about how he thinks about building tech today, and how AI agents are changing the experiences the company creates and measures. He also discussed the importance of security in the AI era. (Qualtrics announced a key security milestone this week.)
In the final segment, I turned the tables on Brad with a game about cloud and AI terms called “Real or Ridiculous” (see below). Listen to the end to play along.
But first, he reminded me about the random case of mistaken identity that brought us together in the first place, back in the day.
Listen to the episode to hear the story, and continue reading for highlights from his comments, edited for context and clarity.
On the unprecedented pace of AI innovation: “My team sends out a summary of what changed in the last seven days every single Monday to the engineering team, because it’s moving that fast. I’ve never seen anything like that.”
How AI is transforming customer feedback: “Worldwide right now, when someone starts a survey, the completion rate is 75%. … With our Gen AI enabled surveys, what we call conversational feedback, it increases to 83%. … When we ask the follow-up question, we get 30 times the number of words back in the second response. And so with generative AI, we’ve been able to increase the amount of data coming back by 10%, and double the quality.”
On AI’s impact on engineering productivity: “What we’re seeing right now is, Cursor is literally generating millions of lines of new code for us. Of course, it’s AI generated, human-reviewed, human approved, human corrected. But 45% of all that code that’s being generated, we’re checking into the product.”
The changing shape of engineering organizations: “I think, as an industry, there will be fewer entry level [positions]. If a typical pyramid for an engineering organization is that 20%, 15% of their engineers are entry level, that probably goes down by 3% to 5% over the next couple years.”
On trust as the key to AI success: “If you ask yourself the question, who are going to be the organizations that are going to thrive in this world of AI, I would argue it’s going to be the organizations that business leaders trust.”
On using AI to accelerate decision-making: “I’ve spent hours and hours inside of ChatGPT just asking questions such as, ‘Hey, if you were the president of a 1,500-person organization with revenue in the multiple billions … what would you do if you wanted to significantly increase the capabilities and the skill set of the engineering team in AI? … It may not have all the answers, but boy, it puts things in your mind, it gives you ideas.”
And here’s the quiz that I gave Brad in the final segment, which I created with help from ChatGPT, Claude and Gemini. Listen to the show to hear the answers.
Real or Ridiculous?
Are these cloud and AI buzzwords real industry terms, or just expertly crafted nonsense?
- Quantum Cloud Orchestration: A system that manages and schedules workloads for quantum computers hosted in the cloud.
- Federated Learning: A machine learning approach where a shared model is trained across multiple decentralized devices or servers, keeping data local.
- Cognitive Load Balancing: An advanced load balancing technique that uses AI to predict and distribute network traffic based on the cognitive state of individual users.
- Serverless AI Inference: Running AI model predictions without managing the underlying servers, scaling automatically based on demand.
- Hyper-Personalized Edge AI Microservices: A highly granular AI architecture that delivers customized intelligent services directly on edge devices, tailored to individual user preferences in real time.
- Sentient Cloud Nexus: A fully autonomous cloud infrastructure capable of self-awareness and independent decision-making, optimizing its own operations without human intervention.
- Emotion-Centric Journey Taxonomy: A proprietary framework for categorizing user sentiment patterns across multi-channel experience touchpoints, enabling micro-adjustments to brand resonance in real time.
Real or ridiculous? Listen to the final segment to play along and hear the answers.
Subscribe to GeekWire in Apple Podcasts, Spotify, or wherever you listen.
Audio editing by Curt Milton.
