Wovenware, a Puerto Rico-based nearshore provider of smart software solutions, announced it has launched a practice specifically focused on the development of Artificial Intelligence (AI) solutions, including chatbots, predictive analytics, machine learning and deep learning applications.
By leveraging its expanded team of data scientists, advanced AI server, private crowd and best practices, Wovenware is enabling organizations to gain greater business value by automating critical tasks and using data-driven insights to make better business decisions, company executives said.
“While most companies recognize the unprecedented benefits that AI-based solutions provide, the complexity of deep-learning algorithms, the shortage of skilled data scientists and the need for specialized GPU-based servers and workstations make it difficult for them to transition to AI-based computing on their own,” said Christian González, CEO and co-founder, Wovenware.
“Our focused AI practice is designed to help companies accelerate their AI transformation initiatives and begin reaping the competitive advantages immediately,” he said.
AI practice spans full lifecycle of machine learning services
Wovenware’s new AI practice provides the full lifecycle of machine learning services — from preparing training data to deploying deep learning algorithms, executives explained.
Led by Leslie de Jesús, data scientist and Innovation Director, Wovenware has appointed a team of data scientists to work alongside its more than 50 software engineers to implement key AI-based initiatives for clients in industries from government and manufacturing to healthcare, transportation and more.
In addition to its highly focused expertise, the team will draw upon best practices gathered from its successful AI-based projects.
Some of the projects include helping a medical device manufacturer predict the likelihood of failed components; developing deep learning algorithms that enable a government agency to identify specific objects on a map; or creating a chatbot so that a university can better respond to inbound student questions.
Dedicated private crowd
Sorting through vast amounts of data and images, labeling it and cleaning it are critical to the development of accurate algorithms. Wovenware is one of the first companies to employ a dedicated private crowd, or group of data specialists, to accomplish this task and expedite the process of preparing the large amount of clean and accurate data that AI apps require.
The team of more than 50 Wovenware data specialists that comprise the private crowd understand the nuances of how algorithms are taught and how smart apps can learn, to provide extremely accurate and precise identification of objects, images and data sets to ensure the highest quality algorithms, the company explained.
GPU system for deep learning
Wovenware partnered with systems integrator, Microway, to design an advanced, purpose-built supercomputer that provides the GPU capacity required for today’s deep-learning applications, they said.
“We worked closely with Wovenware to design a GPU-based system that that will allow it to rapidly train extremely large deep learning datasets — a process that could take hundreds of hours on a CPU-based system,” said Stephen Fried, president of Microway.
The Wovenware team tells us it is achieving these results in as little as two hours on the new GPU-accelerated deployment. We’re thrilled that Microway’s factory integration of deep learning software is helping Wovenware achieve these results quickly and efficiently,” he said.