From Chat to
Orchestration
Vu Agency was incorporated on 7 June 2024. Since then, AI went from a thing you type questions into to a thing that orchestrates your work.
Vu Agency was incorporated on 7 June 2024. It does not sound that long ago. In AI terms, it is another era entirely. When we started, most businesses still treated AI as a chatbot. Over the following year, mainstream products started to think for longer, use tools and carry out multi-step tasks. OpenAI introduced o1, Anthropic launched computer use in public beta, Google started talking about an "agentic era", and GPT-5 later pulled fast answers and deeper reasoning into one system.
Alex and I had been seeing that change up close before we set the company up. We were both using AI every day and noticing the same thing: work that used to take a week was getting done in a day. At the same time, most businesses we met were barely using it at all. They would mention ChatGPT, say it helped tidy emails, and that was the whole strategy. That gap is why we started Vu.
We have had a couple of encouraging signs since then. Vu Agency won Solihull Breakthrough Business 2025, and this year we were shortlisted for Digital Revolutionary at the Greater Birmingham Chambers of Commerce Awards. We did not expect either, but they were useful signs that the work was landing.
We built what we needed. Then we realised others needed it too.
We redesigned this site with 102.ai, which generates three redesign options in under 102 minutes. We used it on our own site first, put the three directions to a vote, and the version on this page is the one people chose.
The products tell the story better than any deck. Project Quote AI grew out of the usual mess of discovery calls and vague briefs. It lets someone describe a software project in plain English and get a specification, mockup, price and timeline in minutes. 102.ai came from the same instinct. So did Raq.com, which puts more than 20 tools we kept needing into one place, from AI chat and documents to forms, transcription and e-signatures.
None of that started with a grand product plan. We built tools we wanted to use ourselves, then realised other businesses needed the same things.
Different sectors, same idea: remove the slow parts of the workflow.
For a Birmingham events venue, we built a booking and operations platform with a live pipeline, revenue forecasts and an AI assistant that handles enquiries overnight. Before, their team was manually replying to every email and losing leads after hours. Now the AI handles the first conversation, and a human only steps in when the booking is ready to confirm.
For a recruitment company, we built an assessment platform that grades questionnaires, syncs results into the CRM and cuts the admin between application and shortlist. Their consultants went from spending half a day on paperwork to reviewing a finished shortlist.
For a financial advisory firm, we built a protection needs analysis tool that calculates client exposure and produces branded reports for client meetings. FCA Consumer Duty compliance built in from day one.
Efficiency does not reduce demand. It unlocks it.
I keep coming back to Jevons paradox. When a tool gets dramatically more efficient, demand usually rises rather than falls. That is exactly what AI has looked like in practice. The venue that automated bookings did not cut reception staff. They started hosting more events. The recruitment firm that automated grading did not lay off consultants. They placed more candidates. Clients are not using AI to do the same amount of work with less effort. They are using it to take on work they would previously have delayed, outsourced or dropped altogether.
The value is not in producing one clever answer. It is in connecting AI to the work itself.
Search has already moved that way.
Google said AI Overviews had scaled to 1.5 billion monthly users across 200 countries and territories by May 2025, and publishers have since filed antitrust complaints arguing those summaries damage traffic, readership and revenue. For agencies and businesses, this is one of the biggest commercial changes of the period.
Governance is no longer optional.
The EU AI Act entered into force on 1 August 2024, with the ban on unacceptable-risk systems applying from 2 February 2025. In the UK, the government published its AI Opportunities Action Plan in January 2025 and followed it with a progress update in January 2026. Governance is no longer the bit businesses promise to think about later. It is part of doing the job properly now.
Even with all the noise, most businesses are still early. Pew found in June 2025 that 34% of US adults had used ChatGPT. By October 2025, Pew also found that 21% of US workers said AI handled at least some of their work, while 65% said they did not use it much or at all on the job. The headlines make AI feel universal. The reality is more uneven than that.
That is the gap we care about. Not the novelty of AI, and not the panic around it, but the practical work of changing how a business runs.
The next edge will go to the businesses that actually change how work gets done.
We built Vu in the middle of this shift. 102.ai is live. Raq.com is live. Project Quote AI is live. And our free AI business analysis tool is the fastest way to see what AI means for your specific operations.