Rio Tinto: A Copilot Workshop Built Around Real Safety Data, Not Generic Demos
A Copilot workshop built entirely around Rio Tinto’s own safety data, not generic AI demos.
Chevron’s Instrument Engineering team had genuine, grassroots interest in AI, but interest alone doesn’t translate into people knowing how to use Copilot on their own work, so we built a hands-on workshop for around ten team members entirely around a real instrument specification sheet and working Excel file, covering agent building, email-tone automation, and deep Copilot-in-Excel skills including macro generation.
Instrument engineers deal with dense, structured documents – specification sheets, data tables, compliance references – and a lot of recurring spreadsheet work that could be automated if the team knew how. Copilot licences existed, and the team had real interest in using them properly, sparked by an internal poll that showed strong appetite for a hands-on session. But without training built around their actual files, most teams in this position stay stuck using AI for surface-level tasks and never discover what it could really do for their day-to-day work. Chevron didn’t need a generic Copilot overview. They needed a session where the examples were pulled directly from their own instrument specs and Excel files, so what the team learned on the day was immediately usable back at their desks.
We asked for a sample instrument specification sheet and a regularly-used Excel file ahead of time, then built the entire three-hour, hands-on session around those real artifacts for roughly ten members of the Instrument Engineering team – deliberately scoping out Copilot Studio in favour of simple, immediately usable agents built inside the Copilot app itself, with a short Copilot Studio demonstration at the end.
Chevron’s Instrument Engineering team left the session with working Copilot agents built around their own real files, deeper Excel automation skills including macro generation, and email-tone automation they could use immediately. Luke Marston, department lead, confirmed strong interest specifically in the agent functionality and said it would add real value to the team.
| Industry | Energy |
| Format | Hands-on, laptops-required workshop, ~3 hours |
| Focus Areas | Copilot agent building, email-tone automation, Copilot in Excel (incl. macros) |


If you’re considering AI but aren’t sure where to begin, get in touch.