# The WaveHouse Origin Story

> Or, how some engineering undergrads 2,000+ miles from Silicon Valley accidentally made a database management system.

_2026-08-24 · Jack Woods · tags: story, company_

*No generative AI was used at any point during the creation of this blog post. Written and edited by humans in Michigan.*

When I decided to attend the University of Michigan College of Engineering in early 2022, I seriously thought that I’d be a *rocket scientist* by this point. My plan was to study aerospace engineering then go on to work for one of the new space players, like Rocket Lab, SpaceX, Relativity etc. I was excited by the competitive landscape and the challenges yet to be solved; I honestly thought I had it all figured out. 

Things went in a different direction once I met Eric. 

What follows is the story of WaveHouse, an open source project from Wave RF, from the perspective of… me! Jack Woods, co-founder and CEO of Wave RF. But to understand how we got here, and where we’re going, start at the beginning: 2023.

![Jack giving a thumbs up next to Eric, who is eating a sandwich, outside on a sunny day](./images/wavehouse-origin-story-1.jpg)  
*Me (left) and Eric (right, eating sandwich), Apr. 2024*

## Part 1: The Beginning (Jan. - Aug. 2023)

Eric and I met when he was my lab instructor in my first “real engineering” class during my second semester. We had lots of conversations about open-source software, preferred embedded systems architectures, and technology’s general place in society. We grew a lot closer over the summer after that semester, working together to build a custom tracking system for student-launched weather balloons. The division of labor was simple: Eric did the software, and I did the hardware. 

While Eric and I both officially had the job title of “Research Assistant”, we really spent that summer of 2023 essentially doing product development. We were provided a lab in the basement of the Climate & Space Sciences Research Building (which quickly filled with prototypes, DigiKey orders, and multimeters) where the two of us worked together to make a system that met all our requirements: able to accurately report location from as high as the edge of space, withstand temperatures down to -30°C, and have 3+ hours of battery life. Most days it felt more like a startup environment than a research lab (this may be foreshadowing). 

![A bare green motherboard PCB silkscreened “Balloon Tracker Motherboard”](./images/wavehouse-origin-story-2.png)  
*An unpopulated motherboard PCB for our trackers, Aug. 2023.*   
*This gives me the ick now… there’s SO many things I’d change.*

The project wrapped up as classes started again that fall. One day shortly after classes had started, Eric and I were back down in the basement, just yapping about classes, our lab sections (I’d been hired as a lab instructor to teach alongside Eric), and Michigan football, when Eric shared an idea with me that he’d been mulling over: *what if the low-power and long-range radio tech we’d used in the weather balloons could be repurposed to text during football games at the Big House?*

## Part 2: The Birth of Towerless Texting (Aug. 2023 - Feb. 2024)

Like Eric, I recognized the problem, potential applicability of our background, and commercial potential– so I was sold, immediately. For the uninitiated: Michigan Stadium (better known as the Big House) has TERRIBLE cell service. The nearby telecomm infrastructure (or lack thereof) is consistently overloaded by the 100,000+ people in attendance. If you lose your group, want to send a picture to mom and dad, or otherwise need your phone to be Internet-connected: good luck. 

Eric mentioned that he was taking an entrepreneurship class that semester, and that he could use the project for the class (this becomes important later). I told him I’d love to be involved, and that I could provide any hardware development support he needed. Throughout that semester, Eric funded development himself, and many afternoons were spent refining prototypes and designs.

Early on, development was slow. We were underequipped, underfunded, and trying to outsmart billion-dollar telecommunications companies using tiny cheap radios. But, things started looking up in February of 2024 when Eric told me that a professor of his from last semester (remember that entrepreneurship class?) had offered to give him a loan to really pursue this idea. It wasn’t a *ton* of money (though it felt like it was at the time) but regardless, now we were cooking with gas. 

## Part 3: The Birth of Wave RF (March 2024 - June 2025)

Knowing we had to incorporate, we eventually landed on the name Wave RF for this new venture, and WaveRF, Corp. was incorporated at the end of March 2024– a couple of weeks before my 20th birthday. 

Recognizing that we couldn’t keep working out of Eric’s cramped apartment, we began renting office space in July. Suddenly it felt real. We had business cards, branded shirts, and pitch meetings. Everyone we pitched early on came back and said “not yet”, which is how investors say no to you politely while leaving the door open for themselves down the road. 

Nevertheless, development continued. An underwhelming alpha test in the fall of 2024 was a tough setback (we learned a great lesson on the importance of battery undervoltage protection!) but we continued onward, watching the prototype version number slowly climb up through the 10’s and into the 20’s. 

By mid-2025, we were running on fumes. The professor’s money was almost gone and we had little to show for it. Eric and I faced the hard reality that if we didn’t make something happen by the end of the summer, we’d have to pack it up and go home. 

I’m not joking when I say that the very next day after Eric and I sat down and agreed that this summer was make or break, we got an email from [1517 Fund](https://www.1517fund.com/). We’d applied to their Flux Capacitor program earlier that spring, and got invited to interview with them, which had gone fairly well. **That email said they’d like to offer us a $100,000 investment to pursue this.** There are three things I will never forget about this deal: the email notification popping up on my laptop, signing the contract, and **just how close we cut it.** **When their check cleared, we had $187 left in the company account.** 

## Part 4: The Evolution of Wave (July 2025 - Present)

 Things didn’t remain sunshine and rainbows, though. Right as we prepared to order our first production run of the device we’d grown to call the Wave Clip (a keychain that lets you text over radio waves) the federal government put tariffs on, like, everything. Our costs jumped 30-40%, nobody knew whether the rates would stay still or keep moving, and this overall completely broke the unit economics for Clip. 

![A 3D render of the Wave Clip: a blue keychain enclosure with the Wave RF wave mark and a single button](./images/wavehouse-origin-story-3.png)  
*Wave Clip Alpha Rendering - Aug. 2024*

I’ll be honest, this one sucked. We’d had setbacks before, but at least those felt solvable and in our control to some extent. This was just a nonconsensual macroeconomic roller coaster. Focus floated around for a while, exploring enterprise solutions (where the unit economics still actually made sense), exploring software-only solutions (where the unit economics were fundamentally different), and everything in between. 

At this point, we’d begun building an IoT solution based on our previous networking/connectivity experience when we realized that we needed [ClickHouse](/managed-clickhouse). And ClickHouse needed a lot of boilerplate around it to be remotely useful – something not realistic for a small team like ours, and what we’d come to realize is a larger problem than we thought. It was at this moment that WaveHouse was born.

Realizing this problem was bigger than just us, and that ClickHouse’s DX left a lot to be desired, we decided to build a single application that is deployed alongside ClickHouse and does the heavy lifting on auth, realtime streaming, RBAC and RLS. We decided to call this WaveHouse, and to open source it. Why open source? That’s an Eric question, not a Jack question. Maybe he’ll write about it :)

Regardless, that application has now been built. From a DX perspective, (we think) we absolutely hit the mark: the row-level security (RLS) and ability to return ClickHouse data directly from the frontend mean that developers don’t need to concern themselves with spinning up a backend to manage ClickHouse queries requested by the frontend. Our async batching management also means that the ingest path is durable: a server crash or OOM error won’t cause data loss. Since it’s async, it means that WaveHouse will return a 200 as soon as the data hits the ingest path– no waiting for a batch alert or insert. We also built in realtime streaming over SSE, so repeatedly polling the database is unnecessary. When data comes in that matches the stream request, it gets sent immediately – no batching delay included, even though RLS and type validation still apply.  

WaveHouse is open sourced and initial releases have started rolling out. [WaveHouse Cloud](/waitlist), our managed offering, is well underway and launching very soon. Initial test deployments of WaveHouse Cloud have been incredibly successful. 

## Part 5: In Closing (Present - Forever)

So yeah, there’s the rundown: the story of how a couple 20-somethings in Michigan set out to text their friends at college football games, using technology we learned about from building weather balloons, and accidentally made a DBMS for one of the most popular analytical databases in the world. And please, believe me when I say that this **WAS** the summarized version.

It’s been ~3.5 years since Eric and I met on that first day of class. Our journey has been nonlinear, brutally stressful, and also the wildest, most educational, and most unique experience I’ve ever had. While it feels like we’re far from the beginning of the story at this point, somehow I know that the end remains even further away.