← back to blog

Software engineers don't fit in one box anymore

#ai #careers #automation

A new title has shown up: forward deployed engineer. It looks like AI has solved the code problem. Working with AI, code isn’t something we worry about anymore. So what did that lead to? Once knowing how to code stopped mattering much, the value of people who solve problems, entrepreneurial people, went way up.

Companies noticed that productivity went up a lot. Software developers can’t be boxed into software anymore. They’ve started stepping outside their own box and reaching for every solution a company needs.

These people are highly educated, so they can handle problems easily. They have engineering experience, which means a lot of knowledge and practice in solving problems. The only part that takes some time is the domain: finance, accounting, whatever the field is. With AI’s help, even that isn’t very hard.

What companies saw

Companies saw two things happen. First, they started AI transformations inside. A company with 100 employees looked at itself and realized it could maybe produce the same output with 20 or 30 people, and since its costs would drop, its income would go up a lot. That’s why layoffs have been rising so much lately.

Second, as a result, employees are trying to learn more, AI of course, and engineers stand out here. Other employees haven’t adapted to this technical world as much. They know how to use the tools they already have, and learning something new on top of that takes a lot more effort for them. Engineers came to the front because of their problem-solving skills.

I’m sure that in every company, meetings with managers about future career plans have already started. This isn’t only happening to other people. It’s happening in my life right now too, and I’m sure lots of hardworking, entrepreneurial engineers are going through exactly the same thing.

The title problem

It doesn’t make sense to call these people software engineers anymore, because they don’t do the software part directly. It got as simple as handing an idea to AI with a few prompts.

So companies want in on this transformation, and they need to give these people a new role. But the system doesn’t move that fast. These people have a lot of knowledge built up, and working out the right role for them based on that isn’t simple. Then there’s the system around it: the institutions every country answers to, the rules companies have to follow, a lot of things. Because of how companies are structured, nobody can just give themselves whatever title they want. Most companies don’t want to lead on this anyway. They aren’t even trying. Their attitude is more like “let someone else figure it out, then we’ll take it and use it inside”.

So right now, companies don’t know where to put these people, where exactly they belong in the hierarchy as a title. They’re employees, of course. But they’ve become wildcards that work wherever you put them, and nothing stands in their way. They’ll probably come to fear that too, at some point.

For now, they’ll describe these people as employees who automate work in their own departments and hook workflows up to fully AI-driven processes. That will get a name as well. What I’ve heard so far is forward deployed engineer, and it seems to have started in the US. In Vienna, where I live, I’ve seen only four job postings for it. So the shift is still very slow here and still happening. Vienna is late on this, which is natural when you think about Europe as a whole.

Beyond one company

These employees have unlocked new abilities, and they’ve started going at it like they’re starving. They see the opportunities in their departments, jump on them, and start doing everything. Something else opens up for them too. They won’t be limited to one business, or to the company they work for. They’ll be able to apply these solutions in any field they want, with just the few tools they need. That’s all it takes to set these systems up and offer them to companies as a service.

I’m sure most entrepreneurial engineers who jumped on AI early have already set these systems up on their own computers, and get their whole day’s work done by pressing enter a few times. But those setups don’t carry over to other businesses. Even inside their own company, they’d have to install them on other people’s computers in other departments, or buy software and tools and do all those setups for them. And the tools themselves are developing fast.

What it looks like from the inside

Under this huge shift there’s chaos right now. Names and role definitions aren’t settled. Hierarchies aren’t settled at all. To give an example from my own life: at my current job I’ve taken on the responsibilities of backend, frontend, design, product and infrastructure, and I’m moving most of it forward with close to a 100% success rate. I built workflows for it, and I can run the whole system through those workflows. So inside my department, the IT department, there’s no goal left for me to set. To find my own limits, I’m talking with my company right now about what I could take on in other departments and what’s possible there. I’m sure tens of thousands of people are going through the same thing right now.

Across the whole industry, I think this is a very small minority. Most people do it quietly in the background. The ones doing it openly inside their own companies, with an entrepreneur’s attitude, are putting a lot of effort into this transformation together with their companies, I’m sure. I’m going through a similar process, and in it you can see that all the roles we’ve defined for years are shaking at the foundation. For most company employees, we may be heading somewhere not so good.

Companies are ruthless about automating this work, and in the end people will lose their jobs. Those people will want to get back into the market. They’ll use AI to build up their knowledge and skills and try to get back in, so I think we’ll see an explosion: anxiety and fear about finding work, a lot of unemployed people, and a lot of startups and know-how. It will calm down in a few years and settle into some kind of order. But once intelligence is worth very little because of AI, people will start finding new kinds of value. Changes that shake every economic system are coming with that. We don’t really have a choice except to be hopeful about the future. Being pessimistic doesn’t help anyway.

Where I think this goes

I’m sure the next few years of this transformation will look like this. Companies will probably let their experienced engineers loose in other departments to automate every workflow: the ones who took the initiative with AI and have outgrown their department and their field. Depending on the size of the company and what it does, that could take a year or two, or a few years.

After that, I don’t think these people will stay at those companies. They’ll notice the patterns that repeat, the similar flows that follow each other, and they’ll be able to apply them to other businesses. Two or three years from now they’ll have much stronger AI and much better tools. I’m sure the industry will have settled by then and know what it’s doing.

In the end there will be crowds of people with enough AI experience and tools to automate the digital work of every business in the world. They’ll go to companies, small businesses, government offices and so on, and start automating every process a person touches digitally. In short, we’re entering the age of automation. In digital work, a person’s direct interaction with a computer will come down to talking and writing. Instead of clicking things, scrolling around a screen and turning one piece of information into another, people will check what the system produces and see whether everything looks right. And the people building these systems will be founders, of course. Those founders will go everywhere and set it all up.

I don’t know where we’re going in the end, but from what I can see inside this shift, companies are making this push. And we, the founders of this era, used to be called engineers. Now there’s so much intelligence available to us that we can automate every system, so I don’t think we’ll fit in one box anymore, and I think the biggest potential in the near future is right here. My own ventures show it, and I’ve already started moving on it.

If you’re an engineer right now

Over the next few years, if you’re an engineer who doesn’t know what to do next, and you have some entrepreneurial drive and motivation with the AI you have, the best thing you can do should never, ever be writing a mobile app, a game, or plain software to sell to other businesses. That might feel like it could work in the short term, but in some areas its future isn’t bright. When the cost of software comes down to one person explaining their problem to a computer, the best work for system builders and engineers like us is building the infrastructure that turns existing workflows, all those complex processes, into AI automations. That’s where the value is now. At least for a while. In the long run maybe that job gets taken from us too, but that’s a separate discussion, or a separate blog post.

In the short term, instead of writing new software or another web service, you can write the infrastructure for the things that build automation. Or you can make things that help businesses and companies as they move into the automation age. Or you can offer them the service directly. That would be the best venture you can start.

If you’d rather keep working for a company, my guess is that the only way to keep your current job, and the most effective one, is to stop worrying so much about the future and specialize, like a forward deployed engineer, in automating your company’s internal workflows. Or move into roles that build the infrastructure connecting your company’s systems to other businesses through automation. These roles already have different titles, with all kinds of prefixes and suffixes, and there’s no need to list them all. One thing is clear: to protect your job through this shift, show and prove that you can build automation, and look for roles that fit that.

If you’re at the point where you say “I’m not staying, I want to do this as my own business”, then dropping by businesses and companies and trying to convince them how to automate their workflows looks like the right effort to me. You can found your own company and build an AI startup, of course. But given what it costs, the chaos in the market right now and the size of the investments involved, it looks like the riskiest step to me, because it may only pay off in the very long run. I’m sure brave, very entrepreneurial people will choose that path anyway.

I wanted to share where I think software engineering is heading very soon (the long term, I don’t know) and what you can do about it. I’ll keep sharing my thoughts as things develop.