I'm an independent developer working with Byld. I do two things: forward deployment — working inside your team on your system — and regular development — building the thing that's been on your list for a year. AI has made one person enough for both.
Jared Goldman · Mexico City · working US and EU hours
Most people need one of these. They are different jobs.
I work inside your team.
I join your standups, use your tools and your ticketing, and ship on your cadence. Good when the work is ongoing, when there is a backlog nobody internal has time for, or when you need a senior person for a season rather than a headcount.
See how this works →I build the thing, then hand it back.
A scoped piece of work with a beginning and an end: an internal tool, a product, a system your team has been rebuilding in spreadsheets. You get the code, the documentation and the handover. Good when the problem is clear and the deadline is real.
See how this works →Whichever it is, you are hiring one person — not a team with a markup on it.
Regular development — Gigantes Group
A global executive search firm was buying back its own history: thousands of candidate profiles in the CRM, and recruiters on LinkedIn paying to rediscover them. The exact-match search could not find a Head of Manufacturing Operations for a VP of Plant Engineering mandate, because the words differ even when the person does not. I built the system that fixed it, and three more the desk now runs on daily.
Read the case study →Forward deployment — GroundCrew Labs
I needed something that did not exist, so I built it and I operate it: agents that take work off a board and build it under review, every client isolated with their own agents and their own spend, and a data plane that stays on my own hardware. It is what I mean by forward deployment — I do not just build systems, I run one.
Read the case study →Before this: infrastructure for a regulated crypto exchange, a fund's performance pipeline, and authentication for a public broadcaster.
Three things are true of every system I build, whichever way you hire me, and they are the reason a client lets me near their production.
The most useful thing I have built for a client does nothing at all: it cancels things. The moment a candidate replies, every remaining automated message is stopped. The first contact with a new company is written by a human — not because it couldn't be automated, but because that is where the relationship is won.
Rate limits so nothing gets your accounts restricted. Error queues, so a failed job is a row you can retry rather than a hole in your data. Cost tracking, because metered APIs produce invoices, not just bugs. This is the part that gets skipped and it is why systems survive.
Every outside service sits behind a swappable interface. I document as I go, and I write the handover before the last feature.
I assume production, not demo.
Tell me what you're trying to do and which shape you think you need — or neither, if you're not sure yet. Most work starts with a short conversation about where the time is going and whether this is a problem worth building for. If it isn't, I'll tell you that instead.