Things don't always turn out the way you planned. You try to control your future by making sure you play your cards right and pray the Flop, Turn and River ends up in your favor. Luckily, if done enough times, the results will get more and more reliable! 2024 was one of those years for me.

You might wonder, how did I end up here?

2023 was a bit of a wildcard for me. I finally ended up taking a sabbatical starting November 2022, right as we moved into our new house in Vallentuna, Stockholm, as I once again had burned out from being employed at a company with questionable technical leadership. At this point I had been working full time since I got out of high school (2012) without any break for studying or longer traveling, and it had taken a toll on me. I was diagnosed with Generalized Anxiety Disorder (GAD) and ended up at the lobby of a psychiatric hospital after what felt like HOURS of a continuous panic attack. After getting help from my doctor through SSRIs, I managed to recover over the next 6 months and could once again see how being a normal full time employee slowly sucks all the joy out of my life!

As you might suspect, Embark Studios didn't end up being the place where I could thrive. Although I met amazingly kind and talented people, and surprisingly got my name on the credits screen of The Finals, I look back at that time with pride mixed with some regret. The small team I worked in was collectively punished for "bad performance" during development and ridiculed by certain leaders, which still leaves a bad taste in my mouth for AAA gaming companies. Towards the end I switched team from being a backend engineer working on retention features, to working almost exclusively with C++/AngelScript in Unreal Engine on The Finals, and found some joy in doing technical work that wasn't completely tied to the success of the game.

Molnett

Thankfully, my commitment to my company Molnett and to my co-founders Mikael and Sascha has saved me from much of this anxiety. We started working on Molnett full time at the end of 2023 and it has been an amazingly fun journey since. Starting our own cloud provider, on-premise with our own hardware, has proved to be the hardest and most rewarding project I have ever worked on! At this point I have decided to move on from my previous career as an operator in Cloud and move much deeper in the software stack to learn how to build the foundational technology for Cloud, such as with our project Valv.

So, how has Molnett been going?

During the first 6 months of 2024 we had over a hundred meetings with potential investors with a few of them going far enough for them to consider leading our pre-seed round.
It ended up not going the way we wanted and by June we had to resort to consulting ~2 full times to cover our expenses. At the time I was quite frustrated as we had spent months of human time of no monetary results nor product development, perhaps setting us back a few months. I had also used up essentially all my savings and didn't want to end up with any debt. Refining our pitch for different types of investors and learning the game has been invaluable for me for my professional life and I wouldn't have done it in any other way. But a quick reminder for a future me: don't use up all your liquid funds.

We ended up having to consult full time again to support our lives and families. Initially, this was a painful switch for me due to my previous experience. I ended up working with TopTracer (deploying Kubernetes to 1000+ golf ranges around the world) and SEB (advisor for their first cloud-native deployment on GCP) until just two weeks ago. The experience was essentially the same as at Embark. My closest colleagues were a pleasure to work with and the technical challenge was interesting, but working 9-5 at a company which I have no stake in proved to drain all my energy.

Molnett - Looking forward

I'm very excited to say that I am going back to working more 50% or more on Molnett starting January! The team has invested our available time during the fall re-building much of our foundational tech to support real production-safe clusters.

Here's some of the achievements from the last six months.

We got our first paying customer during early summer and they have been very happy with our services. Our core offering today consists of ephemeral deployments, which clones your long-living environments and gives you temporary ones for e.g. your per-PR deployments (Think Vercel deployments but for fullstack applications). This has proved to be a very valuable offering and I'm excited to expand it to Postgres using Neon during next year.

We can now run our entire stack (essentially a full cloud provider) locally on our laptops! As you might suspect, there's a tonne of services you need to support something like that. We do this by using Bazel + Tilt, hermetically building all our dependencies and starting them all within a Kind cluster locally. It looks something like this:

This means that our development and validation time for e.g. moving from deploying with the API, to running workflows with Temporal, took days, not weeks or months. I'm very impressed with what we have built and I have never seen it work as well as this at any of my previous jobs. I'm sure Mikael will write a post about it during next year, if you are interested!

The last couple of weeks we have been working on re-establishing our platform on a new on-premise cluster and with our intended OS Bottlerocket, which is used by AWS' EKS Anywhere and probably in more places. This means we have to configure and setup all foundational services ourselves, such as our own PKI, linux kernel, etcd, Kubernetes and the rest, and package using the upstream tools. And as a result, we have reliable day-2 operations that would be hard to achieve on e.g. Debian with the small team that we have. When this is done we can finally commit to supporting production-level workloads! I can't wait!

Closing words

In the end, the only thing you can do is to hold yourself accountable for how hard you tried and how you deal with the result. The rest is stuck in "whose feat or fault led to Event X?" land.

And I can now say, with my head held high, that I put in as much effort as I possibly could have during 2024. I learned how to get up from the inevitable falling and I'm heading into 2025 with a mental fortitude I previously didn't have!