Upskilling a finance team to use AI safely and well

The problem

A finance team in Delhi was already using AI tools, but unevenly and without guardrails. Some analysts saved hours; others avoided the tools entirely or pasted sensitive data into the wrong place. Leadership wanted the upside without the data and accuracy risk.

The approach

We ran hands-on training built around the team's real tasks — reconciliations, drafting, summarising long documents — not generic prompt theory. Alongside the sessions we agreed a short, plain set of do/don't rules and a vetted list of approved use cases, so people knew exactly where AI helped and where to keep it out.

The result

Within weeks the whole team was working the same way, with a dozen approved use cases and clear boundaries around sensitive data. Analysts reported saving around three hours a week each, and AI shifted from a risky side-experiment to a normal, governed part of the workflow.