About
We write the code. We run it in production.
Small team, large responsibility. Whoever wrote it runs it. Whoever runs it was in the design conversation. That is how the project stays together after launch.
Values
Four principles we won't compromise.
We don't hand off
What we build, we run. 'Handover' doesn't exist here. If the system crashes on a Wednesday night, the same person who wrote it picks up the phone.
Honest about scope
If something won't ship in time, we say so at scope time, not in the final week. Uncomfortable conversation, better project.
Number, not adjective
Not "it got better" — "−40% cold-start", "+12% conversion". We measure, document, verify. Outcome is not mood.
Transparent by default
Changelog, status, decision trail. You should see not only that something shipped, but why it was built that way and what risks remain.
History
Where we came from.
- 2015
MPTR founded. First clients: native mobile apps.
- 2018
Operations became its own pillar. "We build it and we run it" became the rule.
- 2021
First larger AI project — pre-LLM chatbot with classical NLP.
- 2023
LLM integration into first live systems. Claude- and GPT-based copilots.
- 2025
Own GPU on-prem inference. Llama 4 + Qwen3, data stays in the EU.
Want to meet the team?
Half an hour on Zoom. No commitment, just a clear conversation.