The Puerto Rican software engineering company Wovenware, which recently moved its headquarters in Santurce, is preparing to open a second headquarters in early 2020 in California while exploring the possibility of expanding operations to Washington DC as well.
For the firm - positioned as a provider of Artificial Intelligence (AI) solutions- this move will represent expanding its presence across the Anglo-Saxon market.
Christian González, Wovenware CEO, and founder warned that this growth will in no way imply a reduction of the company´s operation in Puerto Rico but quite the opposite.
"Our service exports model is divided nearly 50 percent abroad and 50 percent here. Every time we grow in exports, we grow locally," Wovenware CEO explained from the company´s new offices.
"The office in California will allow us to connect with the client and support the project," said the company´s co-founder and COO, Carlos Meléndez. And that presence is also anticipated for the federal capital, due to the growing portfolio they have there.
"Next year we will have to hire 50 more employees, 20 of them in the data area. That's the minimum we need according to what we've already sold," Meléndez said about the plan to expand the company´s current staff of 105 professionals.
Wovenware co-founders stressed that the company has not left the tailor-made solutions strategy they envisioned when they started with software development in 2003. Quite the opposite, they have enriched this "boutique" approach by offering even more specialized services that range from design to implementation. And to do so, they have a team data scientists who create the models or algorithms, specialists who clean and transform the information to be adaptable, engineers and software developers, a team focused on interface design and project managers leading the implementation process.
"We offer a complete flow of innovation and we fully deliver," Meléndez said.
Talent and a desire to return
In terms of access to talent, the company´s leadership recalled that going into "deep learning" and other AI areas that didn´t exist in the market represented a major challenge. Fortunately, they stressed, that changed in just three years.
"We have seen that universities are giving it more space and there are more graduated with a solid background. We then bring what's missing," Meléndez said.
They have also expanded the talent search to the mainland because they have seen that Puerto Ricans professionals, particularly in Florida and Texas, are willing to come back to the island. "We have at least two (professionals) back,” he said.
"We reviewed salaries and benefits, and our new headquarters, it has all contributed to receiving more resumes of people who have told us 'I saw the new office and I want to work there,'" González said. "We make a concrete effort to have a world-class workplace."
In turn, González admitted that even with better pay, many professionals in engineering, technology, and data sciences also need to feel that they have not stagnated professionally. This is where the service export model comes into play, because "they can work on very interesting projects at the level of any other place in the mainland."
"We are not using or consuming the API (Application Programming Interfaces) which others have already done. We're creating the models ourselves," González said.
They have also seen how information over the benefits of AI in business has evolved, to the point that "local companies that are already coming to see how we can solve problems with predictive models," Meléndez said.
The way clients arrive to their company has also changed in three years. Their previous marketing strategy focus was to stress that for U.S. companies or federal contractors in sensitive fields, such as defense and national security, it was more reliable and profitable to hire Wovenware because it was a nearshore provider with lower prices.
However, their position in the AI solutions market has left the cost factor in the back seat, since it´s clients who call them looking for very particular services, especially in fields such as automated vision or "computer vision" (CV). This field includes creating automatic learning models that teach machines to accurately identify and classify two- or three-dimensional objects.
And, depending on the goal of that solution, the machine could perform actions based on what it saw without human participation or requiring validation.
This is already being used in biopharmaceutical manufacturing, for example, to identify foreign bodies in vials; in public health to identify pests, or in aerospace and defense to identify and classify objects detected by unmanned aircraft.
According to the study “All Enterprises Need (Computer) Vision", by Forrester Research that was released last June, very few companies have the capabilities to develop CV applications and included Wovenware among that reduced list. The study highlights Wovenware as one of only two companies that already have particular capabilities to train these CV models.