Å Energi Developer Conference 2026

European companies are increasingly advocating for sovereign IT solutions due to rising geopolitical tensions and concerns about dependence on foreign technology providers, particularly from the U.S. and China. Our reliance on U.S. based IT platforms, like Microsoft Azure, is in dire need of revision given the current situation.

We're developers, data scientists, tech people… so why do the same patterns keep showing up in estimates, decisions, and collaboration? Hint: it's not a tooling problem, it's a human one. In this talk I'll share 10 psychological principles from real engineering life, plus a few guardrails you can actually use to build more reliable systems and work better together.

How to start using agentic workflows in GitHub from a clean repository — covering settings, tooling, security and more.

During implementation of Norgespris during the summer of 2025, Glitre Nett wanted to show calculations for how Norgespris would affect customers financially. As Elhub was presenting this through their REST API, we wanted to use the same source of truth as customers would see in Elhub portal when ordering Norgespris. The presentation will explain how to get started with Maskinporten and how you can use public APIs.

How the STTP team built a fan-out system delivering high volume market updates to a growing number of trading algorithms. We will show our initial failed experiments that led to the current architecture and how we implemented it – keeping cost and complexity at a comfortable level while handling 30 000 messages every minute.

Getting better results from coding agents.

AI coding tools have become incredibly capable. Yet the output is often still not as good as it could be. For developers using coding agents in their daily work, the bottleneck is increasingly not the model itself, but the context surrounding it. Inspired by emerging ideas around "context engineering," this talk explores the current popular trends and patterns around how we provide AI systems with the right information at the right time. We'll look at common antipatterns, practical examples, and concrete ways we can improve context engineering across our organization, helping developers get more reliable, consistent, and useful results from coding agents.

We attended QCon conference this spring. We'll talk about our observations about what directions the industry is taking in terms of AI adoption, managing teams and skills, and new thoughts on how we define developers. A talk about the top things we learned that we want to share with you.

Most of us are already using coding assistants — but that's only part of the workflow. How do you use AI for the tasks that slow you down before coding even starts? Julie demonstrates a real app build where AI helps with context gathering (meetings, email, project docs), domain understanding, and decision making — then moves into implementation with GitHub Copilot, exploring the difference between agent-style workflows and working directly in VS Code with Copilot extensions, along with practical techniques to get more out of both.

Expect hands-on examples, tips you can steal tomorrow, and a few honest lessons learned along the way.

In this talk I'll be giving an overview of Erlang/OTP – what it is, how it works, and how we use it at Entelios AG for our systems, and crucially, what makes it weird.

Bad things happen. Bugs, downtime, incidents. From the wreckage we try to understand what happened and how to prevent it from happening again – mostly through a post-incident review, or "postmortem". In this talk, we will explore this process and lay out some tips to help us learn better from incidents, and crucially, avoid blame.

This project shows the learning process in practice – from the individual employee to the entire department – with concrete examples throughout.

Å Energi and society at large rely heavily on open and free software. This brings benefits such as lower costs, flexibility, and transparency, but also challenges related to responsibility, maintenance, and licensing. I will show how I, as a Data Scientist, leverage open-source tools in practice.

How HydMet uses AI to enable faster development, scalable solutions, and new possibilities by moving from Excel-based tools to Python web applications.

This talk will demonstrate a real-world use case where a GitHub Copilot agent autonomously reverse-engineered an energy market API, built a CLI tool, and iteratively refined a model for day-ahead wind bid adaptation in the Nordic power market — driven entirely by natural language.

The presentation shows how GitHub Copilot was used to help migrate a large legacy .NET application into a modern cloud-based solution. Dag will show how to do a structured setup, guided execution, and create a repeatable pattern. The real value comes from combining agent guidance, project context, and engineering discipline — not just generating code faster.

Using an agent with prompts, instructions and skills for refactoring Bicep code to Terraform. Saving time, better quality, more structure and zero errors.

How do we persist one million time series per day while keeping performance and cost under control. We'll look at the business cases we're solving, technical solutions we are using and considering, and how we plan to evolve this in the future.

Copilot has evolved from suggesting code to executing complex tasks through agentic workflows. In this talk, I'll show how to set up advanced workflows, what they can actually do, and where the limits are. You'll get a live demo, practical use cases — and some honest lessons on what works (and what definitely doesn't).

First try at Vibe coding. Lessons learned.

A short primer in OPC/UA and why it is an important toolbox for automation and bridging the gap to IT.