- Eino provides a network design and monitoring platform built for the age of AI
- The platform leverages agentic AI and 3D digital twin technologies
- Born out of Cornell Tech, the startup is exploring the use of quadrupeds for network monitoring
Eino is a relatively new startup that wants to change the way RF engineers design networks.
The company today is launching its Agentic Network Observability platform, which targets enterprises with multiple network technologies and mission-critical use cases.
What makes Eino’s agentic solution stand out is its 3D digital twin approach to deliver insights for almost any wireless networking technology, deployed together or separately. Cellular, Wi-Fi, distributed antenna systems (DAS) are all covered, with IoT coming “soon.”
From academia to Eino
The two founders of Eino come with academic backgrounds. Payman Samadi, the CEO, studied photonics and worked at Columbia University for four years, researching ways to design networks for the future. James Sayre, who is CTO, has a Ph.D. in astrophysics and worked on telescope data before joining CenturyLink as a senior data scientist.
In 2018, they developed Eino in an incubator program at Cornell Tech, initially focusing on spectrum optimization tools for telcos. But they pivoted in 2024 to the agentic design platform because they saw AI coming and knew network design and monitoring were going to need to change to keep pace with everything.
Things quickly gained steam, with Eino acquiring about 40 customers in 2025. Current customers include an unnamed Tier 1 wireless operator, as well as companies like NTT Data, Boingo Wireless and Celona.
Keeping up with AI
“We basically realized that everything that is connected to the networks is changing extremely fast, from the way people use their phones all the way to industrial settings,” Samadi told Fierce.
That means engineers need to change the way networks are designed and monitored – and do it faster.
“What is required for the future is the design has to be done in minutes. At the moment, it’s done in days or weeks with the legacy tools,” Samadi said.
The majority of network monitoring tools don’t have enough information about the physical environment. “They’re receiving some telemetry data, either from the network side or some sort of probes and it’s very hard to do the fully automated AI-based root cause analysis when you don’t have enough information,” he said.
In addition, most tools focus on one type of network – either cellular, Wi-Fi or DAS, for example – and they were built a long time ago. Trying to stick AI on top of a legacy platform doesn’t really work. “We saw this gap, and we realized we were in a very good position because we already have the model and the design and predictive performance,” he said.
They added the observability aspect and put it all together for an indoor and outdoor version of their platform.
Robot dogs in the works
In the not-quite-ready-for-primetime technology category are quadrupeds, aka robot dogs, that could be used to continuously collect data from a network. The idea is for the quadrupeds to be equipped with LiDAR, cameras and Eino’s probes and be remotely controlled so they could be shipped to a customer’s premises to conduct surveys and report back on the network.
The problem with the quadrupeds is the majority are made in China, which makes it hard to source for use in the U.S. on any kind of grand scale.
However, “the great thing about it is that everything happens on the robot, so there’s no data getting out to the cloud. That helps on the security side,” he said.
If it all works out, quadrupeds could be doing site surveys for Eino’s customers in the not-so-distant future.