Elisa's path to zero-touch IT services

The challenge

A service desk running on manual process

Elisa is one of the largest telcos in the Nordics with 6,000+ employees, €2.3 billion in revenue, and also Finland's fourth biggest IT services provider. Their service desk is the front line of its IT business, and often the only direct touchpoint between Elisa and its customers. The volume of requests flowing through it every day is significant. And for years, a large part of every specialist's day quietly disappeared into manual work: typing up call notes, filling in ticket fields, categorizing requests, chasing missing information.

The pressure was coming from two directions at once. Customers were demanding faster, better service. Hiring more people wasn't viable. Asking existing specialists to work harder wasn't either. Meanwhile, the data quality stayed poor because there was simply not enough resources, and the system reflected that.

Business woman at the office wearing a headset

The solution

Zero Touch: AI takes over the documentation layer

In 2022, Elisa committed to a systematic AI strategy built on top of their existing ServiceNow environment.

The approach was built around what they call an automation loop: enrich the data coming in, process it in the background, feed the results back to AI, and let the system keep getting smarter on its own.

In practice, that looked like this:

Phone calls turn into tickets automatically

AI transcribes calls in real time, identifies the issue, creates the ticket, fills in all the fields, and when the matter is resolved closes it and notifies the customer. The specialist just talks and solves. Manual input that used to eat two minutes per call now takes 79% less time. For a high-volume service desk, that adds up fast.

AI reads and routes every email instantly

Incoming emails are read by AI, which identifies the request type, checks whether all required information is present, and if something is missing immediately asks the customer for it. Tickets that used to need 2–5 minutes of human attention now take 88–95% less time.

Knowledge articles that write themselves

Instead of a specialist spending 30 minutes building a knowledge article from scratch, the system aggregates learnings across hundreds of similar tickets, identifies patterns, and generates the article automatically — 93% less time per article. The articles go live immediately. If a user flags one with a thumbs down, it is sent to be reviewed.

Customers can talk to their own data

Instead of requesting custom reports, Elisa's customers now have a conversational interface to explore their own service data. They can ask questions, spot trends, and understand their IT environment without waiting for someone to build a dashboard for them.

students group working on school project together on tablet computer at modern university

The outcome

From administrative overhead to operational intelligence

Specialists who have been using the AI tooling for 18 months or more are now resolving 25% more tickets than colleagues who aren't, without working longer hours. That gain comes entirely from removing manual work, not from pushing people harder.

Overall, 20–30% of all service desk tasks are now fully automated. And Elisa is just getting started. The goal is a fully AI-first first line of support where AI handles every initial customer interaction across voice, email, and chat, and only routes to a human when the situation genuinely requires it.

The time savings are as follows: