In March I gave up on the job search

The job market collapsed first

In March, I gave up on the job search.

Most of early 2026 was cold applications. I had the resume, the skills, the referrals I could scrape together. None of it really worked.

The funnel had collapsed from the candidate side. AI mass-applying tools were doing what they were designed to do, which is push thousands of resumes through every posting within hours of it going live. Recruiters I spoke to talked about queues in the hundreds for a single role, with most applications never reaching a human reader at all.

I was one of those applications, most weeks. I had pages of well-crafted cover letters that no one would ever open. The signal I was trying to send was getting buried in the volume of signal everyone else was trying to send.

It was not a feeling of failure exactly. It was the feeling of pushing on a door that was no longer connected to a hinge.

The pivot

In March I stopped applying.

I built ExplainTheHack instead, as a test. ExplainTheHack is a small product for offensive-security practitioners who need to explain what they do, in their own words, for interviews and writeups. It filled a gap I had felt directly: I could perform the work, and I could not always perform the explanation of the work under pressure. If the gap existed for me, it probably existed for other people.

The question that mattered was not whether ExplainTheHack would be a hit. It was simpler than that. If the job market would not pay me to do the work, could I just ship the work and see if anything came back.

I did not know the answer in March. I do now.

Two months, three countries, seven projects

The answer is yes, slowly.

Aleluya Technologies is a one-person shop. Two months in, there are seven projects in flight across three countries: Spain, the Netherlands, and the Dominican Republic.

ExplainTheHack is one of them, and the only one I can name without checking with someone else first. Habemus is another, a tournaments platform live in Spain. Flamenco is a store in the Netherlands, mid-build. The other four I cannot name yet: an IV-therapy and lab site for a client in the DR, a law firm in Santo Domingo where the work is debt-collection software, helping a client publish her first book in Spain, plus collaborator slots on a book-reading app and some SMB automation tooling.

None of these came from a job board. Not one. Most of them came through people who already knew me, or someone they knew. A friend mentioned my name in a meeting. A former colleague forwarded an email. Someone said “I think I know a guy” to someone else.

Warm referral, every time. Not a professional network. Not investor signal. Not LinkedIn DMs from strangers. Just people who had worked with me, or with someone I had worked with, deciding to mention my name when it was useful.

That is the part of the story that does not show up on a pitch deck. Two months is not a triumph. It is a thin signal, just enough to keep moving.

The production stack

The whole production stack fits in a backpack.

I work on a Macbook Air M1, seven years old, that I have refused to replace because it still does the job. There is also a 500-euro mini-PC I picked up in Spain that runs the always-on services: self-hosted Postiz for social-media scheduling, self-hosted n8n for workflow automation, a few cron jobs and a couple of internal dashboards.

Tailscale connects the laptop to the mini-PC over any network, so the always-on services follow me. Claude is in the loop on almost every workflow I run: drafting, refactoring, writing tests, generating boilerplate, debugging at 11pm when I cannot see the bug anymore.

That is it. No team, no SaaS subscriptions beyond what the work strictly requires, no dedicated office. The constraint is the design.

What I still do not have figured out

Distribution.

Reaching new people is the harder part. Warm referral is the engine right now, and warm referral does not scale on its own. It compounds, slowly, on the visible work. Which means the work itself has to do the reaching, because there is no ad budget and no professional network sitting behind me waiting to amplify.

I am learning this in real time from a client in the Dominican Republic who automates his own distribution, end of his marketing funnel back into the top. Watching the mechanism work for him does not mean I have built it for myself yet.

The other open problem is cross-border billing. Three countries, three sets of tax obligations, three invoicing conventions, currency conversions that eat into margins. None of this is novel, every cross-border freelancer figures it out eventually, and I am still in the figuring-out part.

This is what Thursdays will be. What I am building, what I am learning, what I still do not have figured out.