We're in stealth.

Is your organisation, in any sector and anywhere in the world, trying to work out what frontier AI means for you? We can help.

What we can do for you

1. Understand it: we brief leadership teams on how language models work, from a linguist's view, including their fluency and the errors humans would not make.
2. Teach it: we train teams in any sector and build their capability, from first use of a chatbot to working with agents and tools.
3. Test it: we evaluate models and tools on your own tasks and compare cost against quality, so you choose on the results.
4. Work with agents: we set up working practices for AI agents, with written instructions, clear handoffs between agents and people, and review of what they produce.
5. Build it: we build prototypes and working tools for tasks your team already does, so you can try an idea in use before you commit.
6. Work across languages: we help you use AI across languages and cultures, drawing on translation work and cross-cultural training in some thirty countries.

• Decades across sectors and continents: government, charities, universities, business education and media, in some thirty countries.
• Language at the centre: we come to these models as linguists.
• We work at the frontier every day.
• We test before we advise.

Don't fall off the frontier on either side.

On one side, you fall behind: ignoring AI or waiting it out. On the other, you go over the edge: adopting on hype, without care for the risks.

Tell us what you need.