AI for business: agents, chatbots and LLMs

We tell you where AI really saves work and put it to work with your data.

No hype: measured results.

Productionllama-3.3
Typical response time per request 0.84 s
Response time in the slowest cases per request 2.10 s
Tokens served 7 days 31.4 M
Cost per million GPU included €0.48
Cost for the period €15.07 · 0 5xx errors · 1 scheduled restart

Agents, chatbots and consultancy

If you want a task to run on its own, see automation
  • Consultancy: which processes deserve AI and which don't, in writing
  • A chatbot for your website that answers from your documents
  • Agents that query your ERP or CRM with scoped permissions
  • Answers with sources: the bot says where each fact comes from
  • Hand-off to a person when the question needs it
  • Training so your team can use and review the AI

Models in production (LLMs)

Tell us about your case
  • Model chosen by testing it on your own cases
  • vLLM, quantisation and response caching
  • Dedicated GPU or API, depending on volume
  • Measured response time: typical and in the slowest cases
  • Cost per million tokens and per request
  • Prompt and response logging with agreed retention