Zelos Cloud

Origin Story

The Tool That Didn't Exist

There is a peculiar cruelty at the heart of building safety-critical hardware: the stakes are highest, and the tooling is worst. Software teams get GitHub, Datadog, Figma, Notion, and a thousand other instruments of clarity. Firmware teams get printf and a prayer.

Michael Jaradah, Taylor Keairns, and John Ott lived this reality firsthand - not in a laboratory or a startup sandbox, but inside Tesla Energy, where the systems they were writing code for weighed tens of thousands of pounds and held enough charge to flatten a building. The data sprawled everywhere. Debugging required physical presence. Collaboration meant emailing log files. Testing meant hoping for the best.

When John Ott moved from Tesla to Neuralink - going from industrial batteries to brain implants, from kilowatts to milliwatts, from one kind of mission-critical to another - he found the same problem wearing a different uniform. The hardware changed. The tooling chaos did not.

So they built Zelos Cloud. Not because the opportunity was obvious, but because they had run out of patience waiting for someone else to fix it.

The Platform

One Pipeline to Rule Them All

Zelos Cloud is, at its core, a ruthless simplification. Instead of stitching together five different data pipelines, three visualization tools, and a homegrown test harness held together with shell scripts, engineering teams get one platform that does all of it - and keeps talking to hardware they already have.

The architecture is built for the edge - where cloud connectivity is intermittent, latency is unforgiving, and a missed data point could mean a missed failure. SDKs in three languages cover the major firmware stacks:

That is not a casual SDK list. Rust is the language of safety-critical embedded systems. Python is the language of fast iteration and test scripting. Go is the language of infrastructure at scale. Together, they represent every layer of a modern hardware stack.

The Team

Three Engineers, Two Companies, One Very Specific Problem

The founding story of Zelos Cloud is not one of a pivot or a market thesis discovered in a spreadsheet. It is the story of three engineers who met as interns at Tesla Energy, spent years building firmware for industrial-scale systems, and eventually compared notes on how broken the tooling was. The company emerged from that conversation.

This background matters because Zelos Cloud is not a product built by people who studied the firmware market - it is a product built by people who were frustrated enough to quit their jobs and solve it themselves. That distinction tends to produce different software.

Why It Matters

The Infrastructure Layer That Hardware Teams Have Always Needed

Mission-critical systems are not a niche. They are the backbone of the physical world - the electric vehicles people commute in, the grid-scale batteries that stabilize power infrastructure, the medical devices implanted in human bodies, the industrial systems that keep manufacturing running. The firmware that runs these systems is, in many cases, the difference between function and failure.

The target verticals map cleanly onto where the founders spent their careers: automotive and EV, industrial energy systems, medical devices, defense and aerospace. Each of these industries has spent decades writing their own internal tooling - expensive, brittle, and impossible to maintain. Zelos Cloud offers something they cannot build themselves for less than seven figures and years of effort: a purpose-built, maintained, and evolving platform.

The revenue trajectory - from a standing start in mid-2023 to an estimated $440K-$600K ARR based on third-party data - suggests the problem is real and the market is paying.

Technical Foundation

Depth That Shows on the Spec Sheet

The technical credibility of Zelos Cloud is not accidental. When your founders have written production firmware for systems where failure has physical consequences, certain things become non-negotiable: minimal overhead, protocol flexibility, real-time fidelity, and security by default.

The platform architecture reflects this. Data capture runs close to the hardware, with low-overhead SDKs designed for embedded environments where CPU cycles and memory are not abundant. The transport layer handles intermittent connectivity - a reality for edge and industrial deployments that cloud-native platforms often wave away.

The integration with GitHub Actions is particularly pointed: it signals that Zelos Cloud is not asking firmware teams to abandon their existing development workflow. It is inserting itself into the workflow firmware teams already use, which is how infrastructure companies tend to grow - through adoption, not replacement.

Company Timeline

From Tesla Interns to YC-Backed Platform

The Details Worth Knowing

Footnotes That Actually Matter

They met as interns. All three co-founders crossed paths during internships at Tesla Energy. It is the kind of formative experience that either bonds you or burns you - in this case, it built a founding team.

The full stack, literally. John Ott has written production code at every abstraction level from bare-metal assembly to Kubernetes cluster management. Most engineers specialize. He apparently did not get the memo.

The branding evolution is meaningful. Starting as "the first cloud platform for firmware development" and widening to "data platform for mission-critical systems" is not marketing drift - it is a deliberate expansion of the total addressable market while keeping the core product sharp.

Small team, real revenue. Estimated ARR in the hundreds of thousands with three to four employees means each person is generating disproportionate value - which is the startup metric that actually matters.

Find Zelos Cloud

Go Deeper

Explore the platform, read the documentation, and follow the company's progress.