solita.fi

Solita is a tech, strategy and design company, and after 30 years and over 2,200 professionals across nine countries, we’ve learned that the way you lead matters at least as much as the strategy you follow. So we spent time putting into words what we think about leadership. Not as an abstract management philosophy, but as a lived commitment. The result is the Solita Leadership Manifesto: a statement of how we want to lead, why it matters, and what we’re committed to.

In May, I attended the Microsoft Partner Architect Seminar organised by Arrow and chose the solution track “Development of Agentic AI Solutions”. The day’s message can be distilled into a single sentence: Before, we automated from code to production. Now we automate from idea to value, through agentic orchestration. This isn’t just buzzwords.

If you want to understand our culture, spend a day in our Slack. We obviously use it for work, but it’s also a place where you’ll find memes flying, emoji reactions piling up, and our CEO kicking off major company announcements with a picture of a potato. Yeah, you read that right. A potato. Slack isn’t just another tool here; it’s the tool. It’s how we stay connected across countries, offices, projects, and time zones. But what makes it work is how Solitans use it. The vibe is casual but caring.

I want to be at the forefront of AI development I work as an AI Engineer, and my role is very broad by nature. AI is evolving so fast that no two projects are exactly alike, which is one of the things I enjoy the most about this work. My background is especially in AI agents and generative AI, but recently I’ve also worked with Edge AI, DevSec-related projects and different types of AI-assisted development processes.

For service-intensive Original Equipment Manufacturers (OEMs) building products like construction machinery, heavy industrial equipment, or advanced B2B technology solutions, the core challenge is maintaining visibility over machinery you built but no longer physically control. To protect service margins, hit strict availability SLAs, and deploy new data-driven services, you need clear insights into how your equipment is used and how it performs in the field.

Right now, every manufacturing leader is facing the same strict mandates: cut operational costs, improve energy efficiency, hit aggressive corporate manufacturing decarbonisation targets, and manage the volatility and uncertainty in energy markets. But there is a massive disconnect between boardroom objectives and the reality of factory floor operations at production sites.

Growing into the role through hands-on experience I work as a Software Developer at our Lifecycle team, focusing on maintaining and developing existing systems. I joined as an intern a couple of years ago, continued in a student position alongside my studies, and moved into a full-time role right after graduating. That path gave me the chance to show how I can learn and contribute, which was a great way to start my career. Currently, I’m working on two projects.

Meet our People of Solita. We have over 2,200 employees in various locations with different skills and backgrounds – each with a story to share about what they do and what inspires them. This is the story of Stiig, who works as a Senior Software Developer in Solita’s Lifecycle business area in Denmark. He’s been part of Solita’s team since 2020. I never want to get too far away from coding My true passion is development and code.

Suomalaisen Terveys- ja hyvinvointitoimialan tietojohtaminen on ratkaisevassa murroskohdassa. Dataa on enemmän kuin koskaan, mutta sen arvo ei synny määrästä, vaan siitä, saadaanko tieto luotettavasti osaksi päätöksiä, palveluprosesseja ja arjen työtä. Nykyinen raportti- ja työkalukeskeinen tietotuotanto ei enää riitä toimintaympäristössä, jossa palvelutarve kasvaa, resurssit niukkenevat ja tekoälyn hyödyntäminen edellyttää yhteismitallista, jäljitettävää ja käyttötarkoitukseensa sidottua dataa.

A common assumption is that generative models make predictive models less relevant. The opposite is closer to the truth. A generator that produces billions of novel structures is only as useful as the filters downstream, and those filters are often predictive models. This post goes through the common drug discovery pipeline, the role of machine learning in it, and a closer look at molecular property prediction as a component.