Our content
We use AI as a writing aid for texts and articles. Everything we publish is checked and owned by our team.
Transparency
We build solutions with AI, so it's only fair to say openly how we use it, including in our own work and on this site.
Last updated: 02.11.2026
We use language models (LLMs) for specific tasks: extracting data from documents, classifying and summarizing messages, drafting replies and offers, searching company documents. Each agent only has access to the systems and data it needs.
Important actions (payments, customer messages, accounting entries) go through human approval. Everything an agent does is kept in a log you can review.
We don't tie ourselves to a single provider. We pick the model by task, cost and data requirements: models from Anthropic (Claude), OpenAI, Google or Mistral, or open-source models running on the client's infrastructure when the data must not leave the company.
AI does not make decisions with significant effects on people on its own. In recruitment, for example, it can help read CVs, but the decision always belongs to a person.
Requests sent through the contact form reach our team by email. When we use AI to process them (for example, a summary or a draft reply), we use Claude, by Anthropic. A person always reads and sends the reply. Details in the privacy policy.
If we add an AI assistant to the site, it will introduce itself as AI from the first message.
We use AI as a writing aid for texts and articles. Everything we publish is checked and owned by our team.
If you want to know more about how we use AI, write to us at contact@digitalwap.ro.