From Chip Veteran to AI Infrastructure Founder

The story begins almost 12 years ago, when Tayo Adesanya launched his career working with microchips and AI processors. He spent those years advising large manufacturers on which chips to integrate into their hardware, gaining a front‑row seat to the evolving demands of the AI computing market. Adesanya told TechCrunch that this experience gave him early insight into where the industry was headed, and that launching Lola Vision Systems was a deliberate bet on those future trends.

Source: techcrunch.com

What Lola Vision Systems Builds

In 2024 Adesanya founded Lola Vision Systems, an AI infrastructure company headquartered in Washington, D.C., that creates both software and semiconductor chips for running AI models directly on devices. The company’s core offering is a compiler toolchain that translates AI models—whether custom‑built or sourced from open‑source repositories—into the specific instruction set a target chip can execute. Adesanya describes this translation layer as a massive bottleneck, noting that manually configuring an AI model on new hardware can consume roughly 200 hours before any testing can begin.

Automating the Setup Process

Lola Vision says it has rebuilt that software layer from the ground up and is simultaneously developing its own silicon to further automate the workflow. A client simply supplies its code and the desired AI model; the platform outputs executable instructions for the client’s chip. According to Adesanya, speed is only part of the advantage. Faster setup frees aerospace and other mission‑critical organizations to run more accurate models on their own data while staying within tight power envelopes.

Why Accuracy and Reliability Matter

‘For these customers, accuracy and reliability aren’t nice to have. They determine whether a product passes regulatory review and whether it works reliably in the field,’ Adesanya emphasized. In sectors such as aerospace, defense, and industrial automation, a mis‑identified object or a lagging recognition model can have serious safety and compliance consequences. Lola Vision’s approach aims to eliminate the guesswork that currently forces teams to spend days or weeks getting models to run at all, followed by additional weeks of debugging.

Edge Computing Challenges and the NVIDIA Jetson Alternative

Edge computing—running AI directly on a device like a camera or drone rather than in a remote data center—remains hampered by power and compute constraints. Adesanya pointed out that many teams start with NVIDIA’s Jetson modules or open‑source AI models, only to find that these solutions often break or perform poorly out of the box. The resulting integration effort can consume weeks, and even after models are functional, power consumption frequently blows edge‑computing budgets or the board lacks sufficient compute for medium‑to‑large models, causing recognition lag or misreads.

Traction, Partnerships, and Funding

Lola Vision reports that a dozen corporate customers have signed letters of intent to purchase its chips once they become available, and it already has one signed customer. To accelerate revenue, the company will license its software on existing hardware while awaiting its own silicon. It has also partnered with SCALE, a microelectronics workforce development program, to engage more semiconductor labs. To date, Lola Vision has raised just over $1 million in total funding.

Recognition at TechCrunch Battlefield 200

The startup was selected for this year’s TechCrunch Battlefield 200, a cohort of 200 early‑stage companies chosen for the program. Adesanya recalled that TechCrunch was a favorite publication during his time at Purdue University. After roughly a year of product development and securing its first client, he felt the timing was right to apply for Battlefield and gain broader exposure. He said he looks forward to making meaningful connections, learning about industry developments, and, candidly, to seeing investors write checks.