Lola Vision Systems, an AI infrastructure company founded in 2024, is working to make it easier for companies to run AI models on chips used in devices. The Washington, D.C.-based startup is developing software that translates AI models into instructions that specific chips can execute, while also working on its own semiconductor chips.
The company was founded by Tayo Adesanya, who began his career nearly 12 years ago working with microchips and AI processors. His work involved helping large manufacturers determine which chips to use in their hardware. Adesanya said those years gave him an early view of the direction of demand in the AI computing market.
“Starting Lola Vision Systems was a bet on where the world was headed and what I was seeing,” he said.
Lola Vision’s core product is a software layer that the company describes as a “compiler toolchain.” The software is designed to translate a customer’s code and AI model into instructions that its chip can execute. The model can be custom-built or open source.
According to Adesanya, manually preparing an AI model to run on new hardware can take “roughly 200 hours” before testing can even begin. Lola Vision says it has rebuilt this software layer with the aim of automating more of the process.
“Speed is only part of it,” Adesanya said. He explained that faster setup gives aerospace and “other mission-critical companies” more time to “run more accurate models on their own data, at a lower power.”
“For these customers,” he said, “accuracy and reliability aren’t nice to have. They determine whether a product passes regulatory review and whether it works reliably in the field.”
The startup is also developing its own semiconductor chips as it seeks to provide an alternative to Nvidia’s technology for running AI directly on devices. Many companies currently use Nvidia’s Jetson compact computing modules or open-source AI models for edge AI applications.
Adesanya said these solutions “often break or run poorly out of the box, so teams spend days or weeks getting them to run at all, then even more weeks debugging until the models are usable.”
“Even then,” he continued, “power consumption often blows edge computing budgets, or the board can’t deliver enough compute for the medium to large models the product actually needs to run successfully. This leads to the recognition models lagging behind targets or misreading objects.”
Lola Vision said a dozen corporate customers have signed letters expressing interest in purchasing its chips once they become available, while the company already has one signed customer. It has also partnered with SCALE, a microelectronics workforce development program, to work with additional semiconductor laboratories.
The company plans to generate revenue sooner by licensing its software for use on existing hardware rather than waiting for its own chips to become available. Adesanya said Lola Vision has raised just over $1 million in total funding to date.
Lola Vision has also been selected for the TechCrunch Startup Battlefield 200, a group of 200 startups participating in the programme. Adesanya said he had followed TechCrunch since his student days at Purdue and decided to apply after about a year of developing the product and securing the company’s first customer.
Speaking about the event, Adesanya said he was looking forward to “making meaningful connections and learning as much as I can about what’s happening in and around our space.” He added, “And, to be direct, I’m looking forward to investors writing checks.”
The company is positioning its software and chip development around the growing need to run AI models directly on devices, where computing power, energy consumption, accuracy and reliability can be important factors for customers.
