Nicholas D'Andre

A man, a tunnel, and a camera that finally pays attention

Drive into the Lincoln Tunnel on a Tuesday at rush hour and you are one of roughly 120,000 vehicles funneling through a tube finished in 1937. The walls are tile. The traffic is eternal. And somewhere in that river of brake lights, a piece of software built by Nicholas D'Andre is watching every lane change, flagging the near-miss that almost happened, and quietly counting the emissions nobody asked it to count.

That software is GridMatrix, and D'Andre is its CEO and co-founder. The pitch is almost stubbornly unglamorous: he does not sell cities new hardware. He does not bury sensors in the asphalt or bolt fresh cameras to every gantry. He takes the cameras a transportation agency already paid for - the grainy public feeds already pointed at the road - and runs them through cloud-based machine learning until they start producing the kind of analytics that used to require a research grant. Congestion. Signal performance. Collisions. The almost-collisions that never make a report but predict the ones that will.

It is an idea that sounds obvious right up until you realize almost nobody was doing it. Most American infrastructure is, in the most literal sense, legacy hardware. The roads are old, the wires are older, and the data those systems generate mostly evaporates unwatched. D'Andre's bet is that the smartest thing a city can do is not rebuild, but finally listen.

The origin was a layover, not a lab

GridMatrix did not start with a whiteboard. It started, by D'Andre's own account, on the road. He and his co-founders were traveling constantly for their jobs, and kept noticing the same uncomfortable thing in airports and taxis abroad. "We were traveling a lot for our jobs and saw infrastructure that was light-years ahead of what we had at home," he has said. The gap between what was technically possible and what American streets actually delivered was wide enough to drive a company through.

So in 2021, in San Francisco, they did. The premise was that the United States did not have a sensing problem - it had a software problem. The eyes were already on the road. What was missing was a brain.

From spreadsheets to streetscapes

D'Andre's resume reads like a tour through the operational engine rooms of American business, which is exactly the background you would want for someone trying to wrangle physical infrastructure. He started as an investment analyst in McKinsey & Company's New York office, helping manage the firm's public equities portfolio with custom software, market data and a lot of financial modeling. During his MBA he interned at Google, where he worked on demand forecasting for the launch of the Pixel phone - the unsexy, high-stakes math of guessing how many devices the world will buy.

From there he went to Amazon as a senior inventory manager on its core retail business, using big-data analytics to decide what to buy, how much to hold, and where to send it. Then came roughly three years at Apple as a senior global supply manager, where he oversaw hundreds of millions of dollars in annual spend across more than twenty suppliers of sensors, RF test equipment and mechanical components. He was, in other words, the person negotiating for the very sensors that make modern devices intelligent. Somewhere in that work, the irony became a business plan: the devices were getting smarter every year, and the roads they drove on were not.

The proof was a competition nobody expected him to win

In 2022, the Port Authority of New York and New Jersey went looking for technology that could squeeze more insight out of its existing camera network. More than 150 startups from around the world lined up. GridMatrix won the proof-of-concept. That is how a young company ended up running analytics on some of the most punishing transportation infrastructure in the country - the Holland and Lincoln tunnels, the George Washington Bridge - by January 2023.

The work kept widening. By 2024 the platform had gone live at Port Newark and the Elizabeth-Port Authority Marine Terminal, extending the same idea from highways to the choreography of shipping containers and cargo trucks. The technology now runs across six US cities and into Europe, reading roads, ports, airports, universities and stadiums with the same underlying trick: take the feed, find the pattern, alert the operator before the pattern becomes a headline.

Safety as the headline number

For all the talk of efficiency, D'Andre keeps returning to a more human metric. The point of seeing a near-miss is to prevent the hit. "This information is critical to helping cities achieve their Vision Zero road safety goals, reducing emissions and quantifying the impact of infrastructure investments," he has said. Vision Zero - the goal of eliminating traffic deaths entirely - is the kind of target that sounds aspirational until you have software that can actually measure whether a signal-timing change made an intersection safer. GridMatrix's quiet promise is that it can put a number on the thing cities most want and least know how to measure.

"GridMatrix's software puts advanced transportation analytics in reach for cities that previously may not have used them," he has said - a line that doubles as the company's entire thesis. The advanced stuff should not be a luxury good. It should run on the wires you already have.

The operator, not the visionary

D'Andre is not selling flying cars or a moonshot. He holds two patents in applied computer vision and advises Columbia University's Center for Smart Streetscapes and the Intelligent Transportation Society of America's Emerging Technologies Committee, and he turns up regularly at global mobility conferences. But the through-line of his career is unfussy: figure out the real constraint, then build the least dramatic thing that solves it. At Apple it was supply. At Amazon it was inventory. At GridMatrix it is the gap between the data a city generates and the data a city uses.

He carries a Yale MBA and a Pomona College BA, and a habit of pointing at the obvious thing everyone walked past. The cameras were already there. The roads were already old. The only new thing required was someone willing to make the two finally talk to each other.