MagLab is a small engineering studio. We build software that runs a business: internal systems, automation, data platforms, mobile and web apps, and AI that works on a company's own documents and numbers. Clients come to us when spreadsheets, disconnected tools and manual reports start holding the business back.
Every project is led by one Full Stack Builder - a strong engineer who handles the whole cycle: understanding the business and where it is heading, deciding what to build first so it pays off, architecture, design, code, launch and support. You talk to the person who builds, so a change reaches the product within hours or days instead of going through a chain of managers. AI takes over routine execution; experience decides what to hand to it and how to check the result.
How we work: we start with an assessment on your real data - where money and time are lost and what will pay off first. Then we build in stages, each with a written price before it starts, and we keep running what we built.
What we have built:
- Kvasir, our own analytics product. For Suši Boks, a restaurant chain, it brings the POS, bank statements, e-invoices, delivery platforms and the owner's spreadsheets into one model rebuilt every night: 99.99% of bank payments are matched to supplier invoices, and the owner gets yesterday's picture in one morning message.
- Mosena, an AI journal app for iPhone, built end to end: research, product, brand, app, AI and servers.
- A high-load platform of ten microservices with one event stream, and a pipeline that turned a whole research library into a knowledge graph.
- We work with clients anywhere in the world.
MagLab is a small engineering studio. We build software that runs a business: internal systems, automation, data platforms, mobile and web apps, and AI that works on a company's own documents and numbers. Clients come to us when spreadsheets, disconnected tools and manual reports start holding the business back.
Every project is led by one Full Stack Builder - a strong engineer who handles the whole cycle: understanding the business and where it is heading, deciding what to build first so it pays off, architecture, design, code, launch and support. You talk to the person who builds, so a change reaches the product within hours or days instead of going through a chain of managers. AI takes over routine execution; experience decides what to hand to it and how to check the result.
How we work: we start with an assessment on your real data - where money and time are lost and what will pay off first. Then we build in stages, each with a written price before it starts, and we keep running what we built.
What we have built:
- Kvasir, our own analytics product. For Suši Boks, a restaurant chain, it brings the POS, bank statements, e-invoices, delivery platforms and the owner's spreadsheets into one model rebuilt every night: 99.99% of bank payments are matched to supplier invoices, and the owner gets yesterday's picture in one morning message.
- Mosena, an AI journal app for iPhone, built end to end: research, product, brand, app, AI and servers.
- A high-load platform of ten microservices with one event stream, and a pipeline that turned a whole research library into a knowledge graph.
- We work with clients anywhere in the world.
Location and contacts
Major clients
Processes and approach
How do you gather and validate client requirements?
We start with a two-week assessment on your real data and processes: where money and time are lost and what will pay off first. Requirements come out of that assessment, not out of guesses.
How do you ensure alignment with client goals and business strategy?
The Full Stack Builder starts with your strategy - where the business is heading - and decides with you what to build first and how it pays off. Each stage has a written price before it starts, and you continue only if you see the return.
Which software development methodologies do you use (e.g., Agile, Waterfall, Scrum)?
Short iterations: we ship working pieces stage by stage instead of one big release, so a change reaches the product within hours or days.
How do you keep clients and stakeholders updated on project progress?
You talk directly to the person who builds, without account managers in between. Good and bad news come right away.
What quality assurance practices do you follow?
In data systems every number opens the document behind it and results are reconciled against source records. AI features go to users only after passing our checks.
What kind of support or maintenance do you offer after delivery?
We don't hand over a project and disappear: we keep running what we built and support it every month.