12 construction leaders. 25 agents.
Motive International · Construction and MEP · Riyadh
Twelve leaders at a Riyadh construction group went from chatting with ChatGPT to building AI agents for their real jobs. In one bootcamp.
Before
AI meant typing questions into ChatGPT. Inboxes, purchase orders, tenders and meeting minutes all ran by hand. Building something felt like a job for developers.
After
Each leader built agents for their own work. One flags clients waiting days for a reply. One controls project costs. One runs the CEO's day from WhatsApp.
Where they started
Motive is a construction and MEP group in Riyadh. Project directors, planners, finance, and a CEO who wants everyone moving faster.
Like most teams, they used AI as a chat window. Ask a question, copy the answer, move on. The real work still ran by hand: a flood of email every day, purchase orders, tender documents, minutes after every meeting.
The tricky part
Nobody in the room was a developer. On day one, some people spent the first hour just getting the tools installed. One leader ended the first session staring at a "page not found" link.
And they all still had day jobs. One joined the very first session in the middle of a call with a supplier.
What we did together
One week of live sessions, each one building on the last, from talking to Claude properly to running a team of agents. Then everyone picked a "dream project" from their own job, built it, and showed it at Demo Day.
One rule: every demo had to be real work, on their own data. No toy examples.
What changed
Demo Day felt like a different company.
- The leader who hit "page not found" on day one showed a dashboard that sorts his inbox by priority. Live on screen, it flagged a client who had been waiting five days for a reply.
- Bassam, the Project Director, built a cost-control portal with budgets, purchase requests and approvals by email. He built most of it that same morning.
- The CEO watched it and made a decision on the spot: every purchase order now gets coded to a line item, so the company sees profit per item, not just per project.
- The CEO built his own WhatsApp AI secretary. It drafts his emails, books his meetings and finds documents.
- The lead planner built agents that read drawings, BOQs and RFIs. Ask "where is this equipment placed?" and you get a short answer with a link to the exact drawing.
- Finance built a live dashboard on top of their accounting system. As he put it: "I build what I need to work."
Every leader presented a working system at Demo Day. All built in one week, next to their day jobs.
"When Dany told me we would get fully agentified and learn to build our own AI agents in less than 7 days, I thought that was impossible. But we actually achieved it."