RALEIGH — Howso, formerly Diveplane, is continuing to grow amid the boom of AI solutions. The company, which focuses on “explainable” AI, has been busy with new clients including Mastercard and Scanbuy. Most recently, Howso co-founder and CEO Dr. Mike Capps headed to Washington DC to participate in the Senate’s AI Insight Forum.

This activity has led the Howso team to bring on a new CEO, Gaurav Rao. The company announced the hire last week, which will allow Capps more time to focus on national and international AI policy in a new role as chair of the board. Rao, previously an Executive VP and General Manager of AI at AtScale, will focus on scaling the company for a new phase of growth.

I spoke with Rao about his history, his role, and the importance of open AI.

Gaurav Rao, CEO of Howso

Gaurav Rao, CEO of Howso

This interview has been edited for length and clarity.

TechWire: Howso’s focus is on “explainable” AI and the transparency of models. Can you talk about why this is so critical and how you’re adding value to businesses with this kind of model?

Gaurav Rao: Absolutely. So over the last 15 years, AI investment has continued to skyrocket. I think the belief is that by 2027 we’re gonna get to $500 billion with a B in AI spending. Now the challenge is we still see enterprises struggling to get ROI (return on investment) from these investments. So we have high spend, low ROI or low yield. Why is this happening?

We see different types of uses of AI. It could be as simple as “I need a recommendation for the next product I should buy based on market basket analysis.” It could be slightly more integrated into a core process like underwriting. So your AI needs to be able to predict accurately. Or your AI could actually impact society, from predicting failures to saving lives. All of those predictions are using the common sets of ML (machine learning) models, technologies, and libraries.

In each one of those instances, if I’m the business user that’s required to deliver the outcome, I need to have full auditability and transparency into the data and the data pipelines that are making and influencing ML models and those ultimate predictions. And the challenge that we’ve seen in the market today is there’s no consistent way to do that. So for me, when we talk about “explainable” AI, it’s really about helping users whether that’s developers or large enterprise executives get better value from their data by trusting it, making it reliable, making it consistent, so that the predictions that they generate for those outcomes are trustworthy.

‘AI safety is imperative:’ Triangle thought leaders talk artificial intelligence with Senate panel

TW: You came over from AtScale. Can you talk a little bit more about your role there and the kinds of AI that you were working with there?

GR: So AtScale has an AI business within a larger business. The reason why we invested in AI was we saw an opportunity to connect the developers of the world, the data scientists of the world, the ML engineers of the world, and all the incredible outputs that they’re creating. The predictions we’re talking about often live in code, data, and dashboards that are really only accessible and understandable by data scientists. And we recognized that a lot of decision decisions being made by business users are in Tableau and Excel. It would be great if we could connect those outputs from the ML world to where business users are making decisions in their reporting infrastructure.

It was an interesting opportunity to work with business analysts and non-technical users. And what I loved about that experience that I think is going to help here at Howso is when you think about the type of technologies we’re bringing to market, they’re applicable not just to developers, but non-technical users. I think that’s where the power of Howso and the opportunity for Howso is going to manifest going forward.

TW: Howso has picked up a couple of pretty big-name clients, including MasterCard and some government groups, which I think is a really natural fit for Howso’s open source and transparency. What are some of your big strategies for this coming phase of growth?

GR: Great question. You highlighted a growth mindset and with these customers, we’re starting to get more demand for use cases and expansion as customers are starting to use and deploy our technology.

In order to meet those demands, one of the first areas we’re going to start to hire and grow in is R&D. We want to continue to build platform capabilities. And then I think the second area when it comes to hiring is going to be go-to-market expansion. So as we continue to learn and build the muscle memory around the use cases we’re deploying, we recognize there’s an opportunity for us to go into specific verticals and domains like government, retail, finance, insurance, health care, etc. These all become opportunities for us this year.

From Diveplane to Howso: The meaning of AI startup’s name change and its open source future

TW: Sounds exciting. I think 2023 was sort of the year that AI hit public consciousness. It was the year of generative AI. What would you say is going to be the big thing in AI for 2024?

GR: So tapping into my crystal ball I do see some similarities carrying over from 2023. I highlighted the need for ROI. And I think as more large enterprise customers invest in AI, you’re going to see an increased pressure from senior executives to say, “Look, we made the investment. Now what’s the yield? What is it doing to improve my top or my bottom line?” So I think ROI pressures are going to be high this year.

I see an opportunity to develop more open-source models this year. We’re starting to see rapid-pace innovation in the open-source community. And then I think, as we continue to use AI more, and as it makes its way into our everyday life, we’re gonna see increased pressure on responsible use. I think you’re gonna see more people questioning, “Hey, if I’m using this open-source MLM where was this trained from?”

Data sources starting to become closed because that’s their proprietary data and they don’t want LLMs being trained from it. I think all of this is going to increase regulation in the field of AI. Not all of this isn’t necessarily new, but I think 2024 is going to change the acceleration in some of these areas.

TW: There’s going to be plenty to keep an eye on, especially with regulation.

GR: Absolutely. And I love the fact that Mike [Capps] is making time for that work but is still going to be a mentor and plugged into our day-to-day at Howso.