Software development
AI automation for business processes.
We automate the work that eats people's hours: reading documents, writing summaries, answering the same questions again, and moving data from one system to another.
What we automate
The candidates are the tasks that are repetitive, text-heavy and currently done by a person because no rule ever quite covered them.
- Reading incoming documents and pulling the fields out of them.
- Summarising long material into something a person can act on.
- Answering repeat questions from your own documentation.
- Classifying and routing email and tickets to the right person.
- Moving data between systems that were never meant to talk.
From practice
Staxo News is a daily crypto digest in five languages. Every morning the pipeline reads dozens of sources, picks the stories that moved something and rewrites them in plain language. It runs unattended and has done for months.
Staxo Trading Agent is an autonomous agent on Kraken with six signals, self-adjusting weighting, risk gates and a daily report. Both are ours, which means we run the on-call for them rather than reading about the failure modes in someone else's write-up.
How a project runs
We start with one process and one number: how long it takes now. Then we build a narrow version, run it alongside the human for a while and compare. If it does not beat the number, we say so and stop, because a pilot whose only outcome is a bigger pilot is not a result.
Models and tooling
We use commercial models where they are the right tool and self-hosted ones where data cannot leave your infrastructure. The choice depends on the sensitivity of the material and the cost per run, and we work it out before building rather than after.
Where AI is not the answer
If the rule is clear, write the rule. Deterministic code is cheaper, faster and does not need checking. AI earns its place where the input is messy and the judgement is fuzzy, and we would rather tell you that at the first call than sell you a model that a query would have replaced.
From the portfolio
AI in our own products.
Web platform
VisitStaxo News
A daily crypto digest in five languages. Each morning an AI reads dozens of sources, picks the stories that actually moved something and rewrites them in plain language, without jargon or hype. Subscribers get the summary in their inbox and know where the market stands in a few minutes.
Hackathon · AI trading
VisitStaxo Trading Agent
Autonomous AI trading agent running on Kraken. Six signals, self-adjusting weighting, risk gates and daily Telegram reports. Built for the Lablab.ai AI Trading Agents hackathon.
Questions
FAQ
What happens to our data?+
That is decided before anything is built. For sensitive material we use models that run in your infrastructure or under a no-training agreement, and we write down which data goes where. If a supplier's terms do not allow what you need, we pick a different supplier.
What does a pilot cost?+
A pilot is deliberately small, because its job is to answer one question: does this beat the current process. Price follows the scope of that one process, and you get it in writing before we start.
Do we need our own model?+
Almost never. Training your own model is expensive and rarely beats a good commercial one with your documents wired in properly. We would need a strong reason to recommend it.
What happens when the AI gets it wrong?+
It will, so the system is designed around that. Anything that touches money, contracts or customers goes through a person, the model's confidence is visible, and every run is logged so a mistake can be traced rather than guessed at.
Contact
Have an idea? Let's talk.
Tell us about the project. For smaller ones we send an analysis and a quote within a few days. For larger and more complex work we agree the scope of that analysis with you first.