The last mile between a real business and the internet.

focus

  • Knowledge governance & RAG
  • Chat with business data
  • AI customer support & lead capture
  • International B2B websites
  • Multilingual & technical SEO

clients: manufacturers, industrial equipment, distributors, technical products, professional services

Client cases

A Fortune 500 subsidiary, logistics

30k+ support inquiries a year, answers scattered across 5 systems and regional teams. Unified knowledge governance, then source-cited Q&A wired into the ticket system: routine questions self-served, support shifted to complex cases.

A national service group

Same metric, different definitions per team, every number waited on the BI queue. Natural-language querying mapped to agreed metric definitions and validated per figure; management asks directly, data pull went from days to seconds.

A mid-size industrial manufacturer

25 years of engineering behind a ten-year-old product catalog. Rebuilt as an EN/DE international site: specs one click deep, RFQ in two steps; overseas inquiries now arrive with specific model numbers.

Small teams and solo operators

No IT, the owner answering repetitive inquiries by hand. Lightweight site, FAQ assistant, structured intake forms; the recurring questions stopped eating the day.

How I think about the work

Business before technology

Some problems need RAG. Some need a database query. Some need neither. The right answer starts from the business, not from the tool of the month.

AI lowers cost, not standards

AI compresses research, content and production dramatically. What you receive must still be worth what you paid.

Know when to say unknown

A system that admits gaps and escalates to a human is worth more than one that confidently invents answers.

You will never be sold a project you do not need.

Every engagement starts with research into your public presence, products and market, then an honest answer: what to do first, and what not to do.

Sometimes the right first step is a website. Sometimes it is cleaning up knowledge before any AI makes sense. Sometimes it is nothing, and I will say so.

Start with a conversation