Estonia's IT layoffs are freeing up technical people. Companies outside tech should move first

Employment among IT workers under 25 in Estonia has fallen by nearly a quarter in three years, ERR's Aktuaalne kaamera reported on 30 October. In the same report Andre Visse, chairman of Telia's Estonian management board, spoke of 400 to 500 people leaving the sector at once. He expects some unemployment for a while and a fairly quick correction.

The causes differ from company to company. Pipedrive, which ERR says will cut about 70% of its 380 Estonian staff, says it is not using AI to replace people: its chief executive wrote that AI changed how customers find its product. Visse expects Telia to do the same work with fewer and fewer people. ERR gives no figure for how much of the fall among young workers comes from AI.

Urve Mets, who leads research at the skills forecasting programme OSKA, told ERR that IT employment has levelled off over the past three years after a period of fast growth. It has held mostly steady in software development, while the share of specialists under 25 has dropped. Young people, she said, now find it much harder to get a first job than they did three years ago.

The rest of the economy has the gap

Visse pointed to a peculiarity of the Estonian market: IT specialists sit almost entirely inside the IT sector. He sees real room for technically minded people to move into other industries and speed up digitising and automating their processes. Mets makes a similar case to young people, suggesting they look at technology tied to medicine or manufacturing. Demand for technical skills will stay high, she said, but writing code alone will no longer be enough.

That matches what ERR heard from universities and employers. The head of the applied AI working group at Tallinn University of Technology said AI now does code writing and code review well, especially for standard code. A developer's time moves to analysing the problem, handing work to agents and summing up what they produce. The university is shifting its teaching toward analysis, how processes work and how to manage AI.

A manufacturer, a logistics firm or an accounting office rarely needs someone to write software from scratch. It needs someone who can look at how invoices, orders or reports move through the company and automate the dull parts. For years those companies could not compete with tech salaries for that person. This autumn they have a better chance.

Two warnings before you build on AI

The same TalTech expert warned that companies are replacing people with too much enthusiasm and may have to backtrack. He estimated AI could cost ten times as much in two years, since most providers run at a loss today. That is one person's forecast, but it costs nothing to test your plans against it. He also warned that after a couple of years of building only with AI, nobody may understand the code well enough to check or repair it.

What to do on Monday

• ✳Write down the three processes where your staff spend the most hours copying data between systems. That list is the job description
• ✳Decide this month whether you want a technical hire to automate that work, and advertise where people leaving IT companies will see it
• ✳Interview with a real task: give the candidate one of those processes and ask what they would automate, what they would leave to a person and how they would check the result
• ✳Keep an entry-level post open if you can, and pair the junior with someone experienced. You will need people who know your systems in five years
• ✳Rerun the numbers on any AI automation with the model bill at ten times today's level. If the case collapses, keep a cheaper fallback
• ✳Require written records of how each automation works and why it was built that way, so a person can still fix it after its author leaves

Technical people who would have ignored your job ad a year ago are reading job ads now. Decide what you would hire them to fix before someone else in your town does.