As AI advances, Domo is evolving to meet customer needs
Once viewed as an analytics specialist, the vendor's product development plans are now focused on helping users build agents and other applications that can trusted and deployed.
SALT LAKE CITY, Utah – As AI development has become a dominant focus for many enterprises, enabling customers to build AI tools that improve insight generation and operational efficiency has become the main objective for Domo.
Customer concerns are what guides Domo's product development plans, according to Chris Willis, the vendor's chief design officer and futurist. As a result, like many data management and analytics vendors, Domo has responded to surging interest in AI development by adding capabilities aimed at making it easier for customers to build AI applications such as agents and chatbots.
Domo, however, is not alone: Several vendors that once made analytics their main focus have expanded beyond their roots to become broader data platform vendors, providing AI development capabilities within their offerings. Longtime analytics vendor Qlik has made AI development a priority. So have ThoughtSpot and GoodData, among others.
With so many competing vendors focused on the same thing, standing out to attract new customers and grow is a challenge. But Domo is doing its best to distinguish itself by integrating analytics, data and AI governance, and AI development in its newly rebranded AI and Data Products Platform, according to Willis.
Recently during Domo's annual Domopalooza user conference, Willis took time to discuss Domo's approach to product development in an evolving market and how the vendor is working to help customers build cutting-edge AI tools.
In addition, with most AI projects never making it past the pilot stage given how difficult it is to build an application that can be trusted to autonomously take on tasks previously performed by people, Willis spoke about the differences between organizations successfully developing AI applications and those still struggling.
Editor's note: This Q&A has been edited for clarity and conciseness.
Chris Willis
Domo recently unveiled new tools such as an AI library and an MCP server aimed at helping customers develop and deploy agentic AI applications -- what was the impetus for developing the specific features that were introduced during Domopalooza?
Chris Willis: We're very passionate about talking with our customers and getting their feedback. That is one of the most valuable things we have at our disposal. But anyone who has worked in designing anything realizes that you can't ask your customers what you should build for them. But what's come out of [customer interactions] is seeing their frustration, but also their excitement, around what AI can do.
It's important to understand that AI is a technology that has been released on the world that is essentially making us all part of a big experiment. Most other technologies have come out of institutions -- the internet came out of academia, tech and government. AI started as a consumer product as kind of an experiment that has turned into a much bigger experiment that … up to this point did prediction and classification. All of a sudden, we had a paradigm shift where AI is generating things. So, where we're focusing, and why you're seeing the products we're announcing, is an understanding that if AI can generate a lot of answers, you need to make sure those answers are right for your business.
New governance capabilities such as User Impersonation, NAV Configs and the row-level access control in Magic ETL were also introduced this week -- what was the impetus for making improved governance a focus?
Willis: Organizations have a deployment problem. They have lots of ideas. They can see where they want to apply data, people and processes -- automation -- to improve what's going on. But the governance part gets in the way, and it's complicated. We're spent a lot of time building governance into our system, but up until now, it's been seen kind of like a "spinach technology" in that it's kind of good for you. Now what we're seeing is that this so-called spinach technology is essential technology because this is the thing that will limit the growth of AI to the right way.
Does Domo have an underlying ethos -- some idea or principle -- that guides developing new platform tools?
Where we're focusing, and why you're seeing the products we're announcing, is an understanding that if AI can generate a lot of answers, you need to make sure those answers are right for your business.
Chris WillisChief design officer and futurist, Domo
Willis: For sure. It's being customer-centric.
We have always differentiated what we do and how we do it based on understanding what works best for people. It's very human-centered design. If you use Domo, you'll see that there's a picture that signifies accountability and ownership of every object in Domo -- there's a person assigned to it. We've always had a focus on user experience and user interface, and that, I feel, has always been part of our secret sauce. It is understanding the customer, what people do, how they act, what their motivations are and what they're trying to accomplish.
When you look at [other vendors], they're very developer-driven, very engineer-driven. Now, they're all trying to move into the consumer space because AI is allowing people, for the first time, to create highly individualized consumer experiences. People are the x-factor.
AI has been omnipresent for the past three-plus years, so as AI continues to evolve, what does Domo need to do to best serve its existing users that are developing AI tools and maybe even stand out a little bit to attract new ones?
Willis: AI has been a gift for Domo. We're not starting from scratch. We have already built a robust, governed data platform. Where we're spending our time, and what we're bringing to our customers, is the ability to use intelligence to reduce complexity and ultimately allow them to innovate safely.
From a competitive perspective, with other former BI specialists such as Qlik, ThoughtSpot and others also now making AI development a focal point, how can any vendor differentiate itself?
Willis: The way we're differentiating ourselves is by providing what we feel are the most essential missing pieces to the AI equation, which is a completely integrated AI and data platform. We've never been just a BI Platform -- everyone has pitched us as that, but we always thought there was something bigger. Being the platform that we are, an integrated and governed data platform, and adding this intelligent orchestration, closes the loop on what we thought was always missing.
Going forward, being able to create individualized experiences is going to be the big unlock for many organizations. From a user experience and usability perspective, there has been a pretty low bar for enterprise software, so being able to create individualized experiences is one big thing. The other is that organizations suffer from a translation problem. The way that one layer talks to another, or one silo talks to another, is sometimes not ideal. At their core, large language models were built to translate … so what we're able to unlock -- being able to turn sources of truth into content and different kinds of products -- is very powerful.
A governed, secure, scalable intelligence platform is critical.
Despite the best efforts of vendors, customers are still struggling to move AI pilots into production. What more will Domo add to help users more successfully build agents and other AI tools?
Willis: What was displayed [during Domopalooza's keynote address} were customers that were able to leverage AI to essentially vibe code solutions. What's coming is a new context layer. That is going to be the bigger unlock. It's an unlock because we are building a contextual system that essentially allows you to create your organizational brand. That does not exist in any model. It's not written down. It's something that was built over time.
I was talking to someone who runs one of the biggest retail food outlets in the world. He told me they built their own agentic workflow – they didn't do it in Domo yet – and he said that worked amazingly well. Then they ran it again, and they haven't been able to replicate that first result yet. What was missing there was the context to debug these kinds of things.
As customers presented during the keynote address, they often spoke about working with Domo to build AI tools. Since someone from Domo can't be there all the time to help, how ready are organizations to take on developing, deploying and managing agentic systems?
Willis: Most organizations are just kind of feeling their way through. But there are best practices.
It starts with really good leadership. There's a big difference between CEOs who understand and communicate and experiment with AI in the right kind of way and those who don't. The companies that are doing it well are the companies that are open to experimentation, engaging with the technology deeply, and are starting with very low-risk, measurable experiments. The ones who are struggling are the ones who are like, 'Guess what, we're going to replace all of our customer service reps with an AI chatbot.'
There's a difference between tame and wild problems. Organizations are ideally suited to solve tame problems [such as] improving their sign-in rate, or cart-abandonment rate, or customer satisfaction rate. AI is a roll of the dice -- you can put in the same thing numerous times and get a different answer. That's an example of a wild problem. One of the important aspects of solving wild problems is understanding how to do experiments and derive best practices. I don't think a lot of organizations are set up for that. … Innovation is risky, it's slow and has big ramifications if you screw up.
If three years ago the big talk was GenAI and this year, it's agentic AI, what do you think will be the big topic at Domopalooza in a few more years?
I say that because, if I were to extrapolate from where we are -- let's say we have a context engine working, which turns into an intent engine that brings in context that's often overlooked -- for the first time [corporations] will be able to deliver on what they have been designed to do in the first place, which is strategically re-allocate resources to make better decisions that create value. That will result in a new kind of model for making better decisions by understanding how actions affect outcomes. Right now, every organization has been good at turning process into programs, putting in all the edge cases and knowing what all the applications will be. Suddenly, this is shifting to something very different where organizations are now focusing on output and outcomes. That's where I think decision intelligence is going to become critical.
Eric Avidon is a senior news writer for Informa TechTarget and a journalist with more than three decades of experience. He covers analytics and data management.
Chris Willis
