There is a genre of company that is very hard to explain at parties, and Ettifos is squarely in it. You cannot point to a phone in your pocket or an app on your screen. What Ettifos makes is closer to a handshake - the invisible agreement by which a moving car and a traffic light, or two cars, or a car and a toll gantry, decide to exchange a few bytes about where they are and what they are about to do. The category is called V2X, for "vehicle-to-everything," and the "everything" is doing a lot of work in that acronym.
The pitch is easy to state and surprisingly hard to argue with. Cars have gotten smart. They have cameras and radar and, increasingly, the ambition to drive themselves. Roads have not gotten smart at the same pace. The average traffic signal has no idea a car is approaching; the average car has no idea what is around the next corner. Ettifos, founded in 2018 in Seongnam, South Korea, looked at that gap and decided the missing piece was not a better camera but a shared one - a way for vehicles and infrastructure to simply tell each other what they know.
The software-defined bet
Here is where it gets interesting, in the way that infrastructure is quietly interesting. The wireless standards for V2X keep changing. There was DSRC, then LTE-V2X, now 5G-V2X, and there will be something after that. Most V2X hardware is frozen the day it ships: you bolt a box to a pole for one standard, and when the standard moves on, someone has to climb the pole again.
Ettifos' flagship platform, SIRIUS, is built on a "software-defined modem." The claim is that the same box can switch from LTE-V2X to 5G-V2X with a software update rather than a truck roll. If you have ever managed anything deployed at city scale, you understand why this is the entire ballgame. The question a procurement officer should ask is not "what does the spec sheet say today" but "can I upgrade this without a ladder." Ettifos built its company around answering yes.
SIRIUS comes in two flavors, an onboard unit that lives in a vehicle and a roadside unit that lives on infrastructure, and it complies with 3GPP Release 16 - the standards document that governs 5G-V2X sidelink, the direct car-to-car radio link that does not route through a cell tower. Researchers use SIRIUS to test the use cases that make the whole thing worthwhile: forward collision warning, platooning, cooperative perception, automated-driving scenarios. Alongside it, Ettifos ships EDM, a diagnostic tool for watching sidelink performance in real time, because when your product's failure mode is "the car did not get the warning," you want very good instrumentation.
The unglamorous milestones
Deep-tech companies do not win by moving fast and breaking things; a broken thing here is a car that did not brake. They win by showing up to standards meetings and passing conformance tests, which is exactly the sort of work that never trends. In 2024, Ettifos, together with Keysight and Autotalks, completed what was billed as the first 3GPP Release 16 5G-V2X sidelink radio interoperability connection - a Korean startup's platform and an Israeli chipmaker's silicon and an American test giant's equipment all agreeing on a radio protocol at the same time. For a 26-person company, being in that room is the achievement.
Then there is the tolling demo. In 2026, in North Carolina, Ettifos wired its V2X-AIR onboard unit into an Audi and had it communicate with Kapsch TrafficCom's commercially deployed roadside units to complete a toll transaction - low-latency, standards-based, no toll booth. It is a small thing that is also the whole thing: interoperability is the product. A V2X device that only talks to other devices from the same vendor is a walkie-talkie. A V2X device that talks to whatever the highway operator already installed is infrastructure.
Compliance as a moat
For a foreign vendor selling into American roads, "Build America, Buy America" is usually a headache. Ettifos turned it into a strategy. Its V2X-AIR device is now fully BABA compliant, which means U.S. cities can buy it with federal money without filing for a waiver. That is a deeply boring sentence and also a purchase order. In a market where most of the buyers are governments and infrastructure operators, being the box that clears the procurement rules is worth more than a cleverer spec.
The money follows the same logic. Ettifos raised roughly $6 million in a 2024 Series A and closed an approximately $11-to-$12 million Series B in 2025, led by LB Investment with participation from L&S Venture Capital, Kyobo Securities, and KR Ventures. The stated plan is to mass-produce a C-V2X hardware accelerator chip on Samsung Foundry technology, ship a third-generation V2X-AIR for aftermarket and micromobility use, and stand up operations in the U.S. and Europe. It is a Korean startup taping out silicon on a Korean fab to sell into American roads - a nicely closed loop, if it works.
None of this is guaranteed. V2X has a chicken-and-egg problem that has humbled larger companies: cars will not talk until roads listen, and roads will not listen until cars talk. Ettifos' answer is to build both ends and the modem in between, and to be patient enough to let the standards, the demos, and the compliance paperwork accumulate. That is not a heroic strategy. In infrastructure, it is usually the winning one.
Who actually buys this
The customer list for a V2X company is not a consumer app's customer list, and that shapes everything about how Ettifos operates. The buyers are automakers and their tier suppliers, roadside-infrastructure and tolling operators, smart-city and C-ITS programs, defense and research institutions, and university labs that need a standards-compliant platform to test 5G-V2X ideas against. These are procurement-driven, standards-obsessed, patient buyers, which is a polite way of saying the sales cycles are long and the technical bar is unforgiving. It is also why so much of Ettifos' public record is demos and conformance tests rather than launch parties. In this market, a passed interoperability test is the advertisement.
What can someone actually do with the hardware? A research institution can drop a SIRIUS unit into a test vehicle and prototype cooperative perception - cars pooling their sensor data so each one effectively sees around corners it has no line of sight to. A city can fit roadside units to signalized intersections and start broadcasting signal-phase-and-timing data, so approaching vehicles know when a light will change before a human driver could. A highway operator can run free-flow tolling, the way Ettifos did with Audi and Kapsch, skipping the gantry cameras entirely. An aftermarket fleet - delivery vans, micromobility, anything already on the road - can bolt on a V2X-AIR and join the network without waiting a decade for the vehicle fleet to turn over. That last case matters more than it sounds: the fastest path to a critical mass of connected vehicles is not new cars, it is retrofitting the ones that exist.
The team and the temperament
Ettifos is led by CEO Hojun Kim, alongside a compact executive bench - a COO, a chief business officer, two co-CTOs, and a CFO - which is a lot of C-suite titles for a company of roughly 26 people, and tells you the work is technical and cross-functional enough to need specialists at the top. The stated values are the sort of thing every company lists - teamwork, open communication, continuous improvement, quality work, embracing change - but in a business where a firmware bug is a safety event, "quality work" is less a poster and more an operating constraint. The company splits its life across two continents, with R&D anchored in Seongnam and a go-to-market presence near San Jose, which is roughly where the customers and the standards bodies for the U.S. market live.
It is worth being honest about the competitive weather. Ettifos operates in the same neighborhood as Autotalks - which is also, tellingly, a partner - plus Qualcomm's C-V2X chipsets, Cohda Wireless, Commsignia, and a handful of others building onboard units, roadside units, and stacks. Ettifos' distinguishing bet is the software-defined modem and the willingness to own the whole stack, from the RTL IP up through the diagnostic tool. Whether that vertical integration is a durable advantage or simply the cost of playing in a market too immature to have specialized suppliers is the open question. For now, the answer Ettifos is giving is the same one it gives to everything: build the boring middle, pass the tests, and let the connected road arrive on the schedule that infrastructure always arrives on - slowly, and then all at once.