From scattered AI experiments to first workflows in one day.

FINN Partners Israel, a PR and communications team, was using AI here and there. After one workshop day, people were building their own tools.

Case study · Show

The situation

FINN Israel is a sharp, fast-moving PR team. It looks after many clients across media relations, social, content and reporting.

The team already used AI, but each person used it their own way, mostly for basic drafting in ChatGPT and Gemini. There was no shared method and no common language. Nobody had a way to find where AI could save time across the whole team.

Meanwhile the manual work kept growing.

  • Monthly activity reports. Junior staff combed through a month of emails, pulled out press hits, looked up the numbers in Cision and pasted it all into templates. This took dozens of hours every month.
  • Media monitoring. Someone clicked through Google Alerts one by one and removed the irrelevant results by hand.
  • Competitor tracking. It happened now and then, if at all. Visiting LinkedIn pages one at a time took too long to keep up.

The team needed more than a lesson in one tool. They needed a new way to think about their work.

What we did

We designed and ran a custom AI and automation workshop for the full team. Sixteen people came, at every level of seniority. Most called themselves AI beginners.

The hard part was the thinking. We wanted people to stop asking “which AI tool should I use?” and start asking “which piece of my work am I trying to improve?”

We built it around their real work.

Before the day, we mapped where FINN's time went. The content covered their activity reports, their media monitoring and their competitor tracking. Every exercise used a mock client with real regulatory complexity, so people practised inside real limits.

We taught methods that outlast any tool.

The team learned a repeatable way of working: the FINN Prompt Formula, the 95/5 Rule, and Compress, Transform, Automate. The team's main tools are Gemini and NotebookLM, so we taught with those.

We moved from listening to building.

In the second half, teams wrote up real FINN workflows on a planning template and scored each one for how easy it would be to build. They left with a ranked list of what to automate first.

In the room

The day ran from mindset, to prompting, to workflow design, to a live build session. In that session people built their first AI-assisted tools on their own, using no-code platforms.

By the afternoon, people were building.

  • An Account Supervisor started an automated media monitoring tool.
  • A Research Manager began setting up a workflow that formats a daily email.
  • A Managing Partner worked on a thought leadership idea generator fed by live news.

The result

The average recommendation score was 9.1 out of 10. Eight of the ten people who replied gave a 9 or a 10.

Everyone who replied named at least one specific workflow they planned to improve in the following week.

By the time they sent their feedback, several builds were under way:

  • An automated media monitoring tool, started during the workshop.
  • A daily email formatting automation, being tested after the workshop.
  • A competitor monitoring dashboard, in progress.
  • A media contact tracking system for targeted pitching.

One person rated their comfort with AI at 2 out of 10 before the day and 6 out of 10 after it.

The team left with a shortlist of five workflows to build: activity reports, media monitoring, meeting notes that turn into follow-ups, competitor tracking and a thought leadership content engine. They have named people inside the team to lead the work, and a plan sorted into now, next and later.

What they said

“Before I had basic comfort with AI. I knew I could use it but hadn't unlocked the full potential. Now, I feel a lot more confident in using AI to accelerate my skills and learning.”

Daniel Davis, Account Supervisor

“AI could assist in doing way more than I had initially considered.”

Riva Gotlib, Research Manager, FINN Israel

What people took away

Give AI a specific role before you give it a task.

Context is the real skill, more than the prompt.

AI can do far more than basic drafting when you approach it with a method.

Think in workflows, and the choice of tool follows.

“The session got me thinking more critically and strategically about using AI in a smart and sustainable way. The session exceeded my expectations.”

Ellie Hanson, Partner, FINN Israel

What this means for your team

Most AI training teaches tools: how to use ChatGPT, how to write a prompt, where the buttons are. That knowledge ages fast and does not carry over to the next tool.

Our workshops teach workflow thinking. It works with the tools your team uses today and the ones it will use in eighteen months. Your team learns to look at its own work differently and to build systems that fix what slows it down.

If your people use AI one by one, with no shared method, that is the gap we close.

Next: Taking over a live product without switching it off. →

What would you love to hand over?

Tell us your problem
create@vectisbuild.co