Tech Insights 2026 Week 28
Here in Sweden many companies are slowing down for the summer, but the global AI industry keeps on innovating. Last week was one of the busiest in a long time with 18 news items (!). So I tried…
Läs merA delivery-driven approach to agentic AI.
We always start with clear-cut goals of the delivery. A clear scope, measurable acceptance criteria, and production-ready engineering – then, and only then, do we build the agentic solution, model, and software around it.
Always integrated into your core infrastructure and fully owned by you end to end.
Our business 
How TokenTek helps you put AI to work.
Agentic AI solutions
We help you identify processes worth automating, select the right agentic framework and build robust AI agents that fit your operations. And just to make it clear – you own everything, models and code.
Custom transformer models
We have deep expertise in model architectures and foundation model training. We design and train transformer models for generation, outlier detection, classification, analysis, and multi-task learning – enabling a single model to handle multiple objectives efficiently.
Agentic software development
We’ve built entire production systems prompted through AI agents for some of Sweden’s largest companies. Including automated tests, documentation, and production-quality code delivered several times faster than conventional AI-assisted development.
Selected clients
Partners
Network
Career 
12-month AI Engineering Program, built with Hubbau.
We help young data scientists and engineers to build transformer models and agentic solutions from the ground up – so they’re ready to contribute from day one.
For organisations
A pipeline of engineers trained in how AI is actually built, evaluated, and run in production.
For ourselves
A fast track into how real systems and real business work, with high standards and strong mentorship.
How we work
Delivery-driven, with outcomes you can keep.
We’re hands-on, engineering-led, and focused on delivering best-in-class AI consultants and production-ready systems with a handover your team can run.
Discover
Define scope, success criteria, constraints, and guardrails – then prioritise the work that’s worth doing.
Build
Design, implement, evaluate, and deploy to production. Plus observability, tests, and all the bits that keep systems reliable.
Transfer
We promise a handover that sticks. Including documentation, enablement, governance patterns, and a system, and way of working your team can own.
FAQ
Straight answers to common AI questions
What is agentic AI, really?
Agentic AI is when models plan, call up tools and applications, and take actions inside a defined workflow. We deliver with integration, evaluation, guardrails, and a design that holds up in production.
How do you decide what’s worth automating?
We prioritise work that’s possible to deliver and operate – usually repeatable processes with clear inputs/outputs, testable quality, and realistic integration points. Then we validate it quickly with a pilot and explicit acceptance criteria before scaling.
Do you need a custom model?
Not always. If an off-the-shelf, general-purpose LLM meets your quality, privacy and cost constraints, we’ll use it. We build custom models when you need stronger performance, tighter controllability, or have ownership and deployment requirements that demand it.
Who owns the solution?
You do, always. We deliver so you can run, evolve and extend the system, including models, code and documentation.
How do you make AI agents reliable?
We treat them like production systems. That includes evaluation, test suites, observability, clear boundaries for what the agent may (and may not) do, and safe fallbacks when uncertainty is high.
How do you handle security and sensitive data?
We design for your environment and requirements, from data handling to auditability and deployment patterns.
What does a typical engagement look like?
We start with a short discovery phase to define scope, constraints, and acceptance criteria. Then we build and deliver a production-grade pilot, test it, then scale it into operations, and complete the handover so your team can own it end to end.
Tech by Johan
AI in a weekly, practical, opinionated and very useful way
Short takes on what matters in the world of AI. What’s noise, and what’s next – written weekly by our CEO Johan Sanneblad.
Latest from Tech by Johan
Weekly tech insights and analysis
Tech Insights 2026 Week 27
Last week I released an update to my plugin Notebook Navigator for Obsidian, improving startup times up to 10 times for some users. Notebook Navigator is the world’s most popular plugin for the…
Läs mer
Tech Insights 2026 Week 26
For a long time, paying for enterprise software was simple. You bought a seat license, and a human sat in that seat and used the software. Last week that model started to break down as we are shifting…
Läs mer The Einstein tile Learn the fascinating story about the Einstein tile here ↗ 
A symbol for how we think
Our symbol is inspired by the “Einstein tile” – a single shape that can tile the plane without repeating. It’s a reminder that real capability often starts simple with one robust building block, assembled into patterns that scale beyond templates.