Technology products
MSX Invoice is an electronic invoice management solution.
AIOS for business, starting with repetitive work.
MSX Tech packages AI Agents and digital workflows for online shops, service businesses and SMEs. The system starts with repetitive work, connects operating data and gradually suggests next steps so teams can work with more clarity.
Automate repetitive work
Connect operating data
Suggest the next step
MSX Tech
Software developed by MSX Tech, delivery services for businesses and a SKILL Market for AI Agents.
MSX Invoice is an electronic invoice management solution.
AI and software delivery shaped around each company’s own process.
A marketplace for workflow packages (SKILLs) for AI Agents.
Product
Built in house by MSX Tech. MSX Invoice runs on Windows and keeps your passwords and invoices on your own computer.

An electronic invoice management solution
Accountants and businesses that need to keep e-invoices under control, including across several tax codes.
Services
Every service starts from the work and data your company already has.
Design and roll out AI assistants and Agents around how your company actually works, on infrastructure you choose.
Build company websites, sales pages or internal web apps, with room for AI integration where it helps.
Review day-to-day work with your team to find where AI genuinely helps, before investing in tools.
AI chatbots that answer customers and staff on your website, Facebook Messenger, Zalo OA or internal channels, grounded in your own documents.
How we work
Scope and deliverables are agreed before delivery starts.
We talk through your work, your processes and the data you already use.
MSX Tech proposes the scope, approach and deliverables for both sides to agree.
We build and test to the agreed scope, then hand over to your team.
Guidance on use and operation, and help with adjustments after handover, as agreed.
Insights & guides
The latest articles on AI Agents and how to prepare your work for AI.

Infrastructure.
A comprehensive architectural evaluation of AI Agent runtimes: When to execute locally on engineer workstations versus managed cloud enclaves, and how to read compatibility specifications.

Best Practices.
The output quality of an autonomous Agent is decided before execution begins: A practical guide to structuring input datasets, reference schemas, and context constraints.

Architecture.
Explore the modular architecture of AI workflows: Why standalone prompts fail to scale across enterprise teams, and how SKILL packages standardize Agent execution capabilities.
Frequently asked questions
Answers about our product, our services and how we work.
Contact
Choose one task you do by hand each week and start there.
Describe the work you want to handle and the result you need.