Exploring the post-code era

Code is free. Judgment isn't.

AI can build anything now. We explore what that means — building real products, testing new approaches, and sharing what we learn. Not from a distance, but from the inside.

What we explore Read our thinking

10+

Products Shipped

We don't theorize from a distance. We build complete products to understand what AI can actually do.

Daily

Frontier Tracking

We scan the AI landscape every single day. New models, capabilities, implications — tested through building.

Nordic Languages

Everything we build works across the Nordics. Because the most interesting problems often aren't in English.

The hard problems are changing

AI agents now write code, run tests, deploy systems, and iterate faster than any human team. A teenager with Claude can build in a weekend what used to take a team of five three months.

This doesn't mean technology gets simpler. It means the interesting questions change.

What should we build?

How do you know if what you built is actually good?

How do you orchestrate 50 AI systems without chaos?

Who's accountable when agents make the wrong call?

Where we go deep

We explore these areas by building real things. Each project tests a thesis about what AI can do.

See all focus areas

AI-Native Software

Building complete applications by orchestrating AI agents — each with clear context, responsibility, and constraints. Custom tools that fit how a business actually works.

Explore

Voice & Conversational AI

AI phone agents, transcription, and conversational interfaces. The most natural way for humans to interact with AI systems.

Explore

Multi-Agent Orchestration

Tools and patterns for coordinating multiple AI agents. Managing AI the way you manage engineering teams — context, responsibility, outcomes.

Explore

Content Automation

Multi-step AI pipelines for content at scale. Research, writing, translation, publishing — each step with its own specialized agent.

Explore

What We Believe

The companies that win in this era won't be the ones with the most engineers or the best AI tools. They'll be the ones with the best judgment— about what to build, what to trust, and where humans still need to be in the loop.

That's the shift we're tracking.

The shift we're tracking

Building

Yesterday

Hire engineers, manage sprints

Where It's Going

Orchestrate AI agents, validate outcomes

Software

Yesterday

Adapt your process to the SaaS

Where It's Going

Build software that fits your process

Quality

Yesterday

Review code line by line

Where It's Going

Review architecture and trust

Speed

Yesterday

Roadmaps measured in quarters

Where It's Going

Prototypes measured in days

Expertise

Yesterday

Deep in one tech stack

Where It's Going

Fluent across the AI landscape

Mindset

Yesterday

"Can we build it?"

Where It's Going

"Should we build it?"

The Old Model

Where It's Going

Building

Hire engineers, manage sprints

Orchestrate AI agents, validate outcomes

Software

Adapt your process to the SaaS

Build software that fits your process

Quality

Review code line by line

Review architecture and trust

Speed

Roadmaps measured in quarters

Prototypes measured in days

Expertise

Deep in one tech stack

Fluent across the AI landscape

Mindset

"Can we build it?"

"Should we build it?"

Frequently Asked Questions

What is Up North AI?+

A think tank and product studio based in Stockholm. We explore the frontier of AI-built software — by actually building it. Products like CallSia, Helanotan, and Orchestrat.ing are the result. We also publish daily AI news and long-form analysis.

Do you take on client projects?+

Selectively. We're most interested in projects that align with what we're exploring — AI-native software that replaces generic SaaS, voice AI, multi-agent systems. If your challenge is in one of these areas, we should talk.

Who are you?+

Ex C-level from Nordic tech companies who discovered that managing AI agents is a lot like managing engineering teams. The same principles apply: clear context, defined responsibilities, focus on outcomes, and the discipline not to micromanage.

How do you stay so current?+

We scan the AI landscape every single day. New models, new papers, new tools, new implications. Then we test what we find by building real products. It's not until you build the entire pipeline that you truly know if the technology solves all the nuances of a problem.

Can I hire you?+

If what we do resonates with your challenge, reach out. We're selective — we work best on projects that push what's possible and teach us something new. No lengthy sales process. One conversation is usually enough to know if it's a fit.