#StartupsEverywhere: Bentonville, Ark.

#StartupsEverywhere: Vance Reavie, CEO and Founder, Junction AI

This profile is part of #StartupsEverywhere, an ongoing series highlighting startup leaders in ecosystems across the country. This interview has been edited for length, content, and clarity.

Harnessing AI for Efficiency in Retail and More

Vance Reavie is no stranger to entrepreneurship. Using his knowledge from previous ventures, he created an AI platform that intelligently automates manual marketing and merchandising workflows and content generation processes for brands and retailers. We sat down with Vance to discuss the startup ecosystem in Arkansas, AI policy and regulation, grant opportunities, and more. 

Tell us about your background. What led you to Junction AI? 

I started my career in government and politics, but by my early 20s, I realized that wasn't the path I wanted to follow. Technology had captured my interest, so I jumped headfirst into it and became a serial entrepreneur. I've had three startups I've exited. Junction AI is my fourth startup that I started in 2018. I've always been fascinated by the transformative capability of technology and how it can affect government, business, and our personal lives. I've lived around the world. I'm a Canadian Australian and have lived in New Zealand, Brazil, Panama, Costa Rica, and the United States. I've loved seeing how technology and innovation have had their path in each of these countries and what we can learn from each other.

My first companies were all organic builds that had nothing to do with investors. Like every entrepreneur, I would see a problem, think that I can make something that does this better, and improve on that while bootstrapping my way through it. After my first sale, I was able to take some capital, skills, and knowledge I got from the experience and bootstrap my way into the next one. Junction AI has been another new challenge because the costs in AI are much higher than in other areas. With this company I've been learning to scale with the investor-backed route in the States, the center of the AI and VC fund industries.

What is the work you all are doing at Junction AI? 

We optimize and accelerate revenue growth by intelligently automating cumbersome manual processes for marketers, merchandisers, brands, and retailers. For example, if you're a big company selling products on online retailers, you have to manually write product descriptions and titles and tailor them to specific sites. We looked at this and decided to create a platform that focused on ensuring accuracy, meeting compliance requirements, and allowing transparent reasoning. The user can see the reasoning, alter it, and manage that process throughout. They can audit the results, and it takes into account all the factors that need to be integrated into that workflow—hence, Intelligent Automation. 

Ultimately, the company only has to create one template or workflow to produce all the necessary content. Right now, many humans are doing the manual work of editing content for each site. We should move humans to more valuable work that involves nuance and judgment instead of repetitive tasks like cutting and pasting.

What led you to founding your company in Arkansas?

I was living in Austin, a great startup hub in a beautiful city with much to offer. However, we were accepted into an accelerator called FUEL AI/ML, a supply chain program based in Bentonville, Arkansas, home to Walmart, Tyson Foods, and J.B. Hunt just down the road. There are plenty of accelerators out there, many of which teach you how to start a business, but I already knew how to do that. This one, though, focused on navigating the complexities of working within an enterprise—dealing with gatekeepers, compliance, regulations, and all the challenging aspects associated with large enterprises. 

It was an exciting approach because it's an area that I'm fascinated by. It's an area where, if you move the needle a little, it matters a lot in these big companies. The area is growing quickly, and Fuel plays a role, as do other groups that network and try to build opportunities here. The area also benefits significantly from the University of Arkansas and the R&D work they contribute to the community. We even partnered with the University on a National Science Foundation grant, to use AI and forecasting for small business and local produce.

AI policy and regulation are often discussed among policymakers. What would you like policymakers to know about AI and the startup ecosystem? 

There’s a difference between AI companies that use existing models or capabilities and those that create their own. Some companies are productizing something already in existence, taking advantage of underlying capability created by others, like OpenAI. They are often taking these tools and making them consumable by the market they’re targeting. Policymakers often group anyone doing work in AI together whether they are startups or not. For startups, there’s a higher risk when changes are made at a global level—whether it affects large companies like Google or Azure. Policies, and the costs, that impact those larger entities often get passed on to everyone using their services, potentially cutting them off from crucial data or imposing new policies or licensing on how models or data are used. These changes can have wide-reaching effects, the impacts flow down, the large entities can afford these costs but startups cannot.

Then there are startups like us who are building unique AI tools, sometimes with open-source models that we then modify to create our version. We have both made models from scratch and  used third party models which we then use our platform to further develop and customize to our client’s needs.The key difference is we are not just white labeling others, we are creating new innovations.   When we try to sell to an enterprise, we can run into various gates we must pass through, like responsible AI committees or legal departments. . It puts an extra layer of work on us to show we've done our original development and original R&D responsibly, have the guard rails, and have the human component—all before we get to actually making a sale. Larger enterprises risk aversion comes from the regulatory scrutiny they are under, but it all flows downhill to us.

Policymakers need a deeper understanding of this space and the differences in the participants, because there is a general lack of knowledge about AI’s transformative potential, leading to an overemphasis on worst-case disaster scenarios. They need to know the costs and impacts on smaller firms like ours. But mostly I want everyone to sit back and realize that we have many laws about data, privacy, business, and regulation, lets start by seeing what works and is already in place. At the end of the day, AI is a tool. It is a very transformative tool, but it is a tool. 

Policymakers must also stop considering only Silicon Valley when considering AI policy. Here in the heartland, many things are happening, but not many policymakers are paying any attention to AI development in Arkansas. However, it is happening here and I hope they  start listening to us and supporting our needs. 

How have grant opportunities impacted your company? 

We partnered with the University of Arkansas for a convergence accelerator grant from the National Science Foundation (NSF). The idea is that it's not just an R&D grant to the university for academic development. Instead, it encourages the university to partner with private sector companies. The idea is to get the university to partner more with private entities and then commercialize the fantastic research that comes out of it. These academics are not startup people or entrepreneurs in that sense. So they need the experiences that people like me have in the entrepreneurship space. In turn, I benefit greatly from the excellent research they are conducting. 

Are there any local, state, or federal startup issues that should receive more attention from policymakers?
For startups, policy is  not just about regulation. For example, one recent thing that affected us was the R&D tax policy change, specifically that you couldn't expense R&D costs all in one year instead of spreading it across 5 years. Most startups don't know if they're going to be there in five years to be able to take advantage of that, let alone have profits to count it against. So that can hurt when you can't expense it all in one year for a small startup. Policymakers need to understand that there is a big difference between the big entities and the  entrepreneurs that are  building a startup economy. 

Likewise,  immigration is another area where it’s not tech regulation that’s a barrier, but other foundational issues. I'm an immigrant. It is costly and complex to get a visa to work for my company with American employees. I should be able to come here and work legally for my own company. All these fantastic international students in the United States want to be here and contribute, have startups, or work for startups. But they end up having that same issue, having a hard time traveling if they need to and getting work visas. 


All of the information in this profile was accurate at the date and time of publication.

Engine works to ensure that policymakers look for insight from the startup ecosystem when they are considering programs and legislation that affect entrepreneurs. Together, our voice is louder and more effective. Many of our lawmakers do not have first-hand experience with the country's thriving startup ecosystem, so it’s our job to amplify that perspective. To nominate a person, company, or organization to be featured in our #StartupsEverywhere series, email advocacy@engine.is.