Practical AI for real onsite behaviour (33:00)

Introduction
Hi everyone, I’m Jenna Farrell, Co Founder of Scopey Onsite.Scopey Onsite is an AI tool that turns WhatsApp messages and voice notes from site into structured event records for the office.We are backed by Enterprise Ireland and private investment, and we are focused on solving one of the biggest challenges in construction: documenting change in real time.

My BackgroundMy background is in software design, specifically how people use and interact with software in real environments.For ten years I ran a software design agency in Melbourne. We worked on projects such as:
‍
- The Australian Government election management system
- Infrastructure software
- Medtech systems used in aged care homes

In those environments, success came down to usability in real conditions.
- How does a worker twenty metres up a cell tower complete a checklist on a phone in bright sunlight?
- How does a nurse whose first language is not English safely record that a patient received medication?

Software only works if it works in the real world.

The Problem:
Managing Change Outside the Contract.
In my own business, the biggest issue we faced was managing change that happened outside the contract.Projects move fast.

- Clients change their minds.
- Unexpected issues arise.
- Deadlines are tight.
- Documentation always comes last.
- We were constantly doing extra work for free because changes were not documented properly.

On average, we were losing around fifteen percent of revenue per project.

This was not just a project management issue. It sat in a grey area between communication, contract compliance and margin protection.The Same Problem in ConstructionWhen I started speaking with construction companies, I realised the problem is almost identical, but with much higher stakes.Margins are thinnerProjects are larger and more complexContracts are denseUndocumented work can have serious commercial consequences.

At conferences I often hear, “We need to make data driven decisions.”But where does that data actually live?

A large amount of critical project information lives in WhatsApp.The Documentation GapContractors tell me:“We have processes, but staff are not adopting them.”
‍
“We use the client system, but we do not have one for ourselves.”Each year, significant project value is lost to undocumented variations, not because of bad work, but because the work was not captured clearly, fast enough, or in the correct contractual format. Teams end up scrambling to reconstruct events later: Linking cause and effectSearching for photos buried in message threads. Trying to confirm when a notice was dueIt becomes messy very quickly.A nd if you miss a deadline, you can be time barred. Even valid entitlement can be lost if notice is late or incorrectly issued. Contracts often contain strict notice periods. The pressure to manage work and document every change accurately is enormous.

The Human Reality on SiteCan you remember exactly what you were doing yesterday at 11:48 a.m.? Most of us cannot. Now imagine Johnny, a supervisor on a busy data centre project.By 11:00 a.m. he is dealing with:

- A late delivery
- A crew member calling in sick
- Another trade damaging completed work
- Delayed scaffolding
- A client change to spacing

He is constantly moving, answering calls, responding on WhatsApp, taking photos. He does not have time to stop and open a laptop to write a formal report. By the time he does, hours later, the details are blurred. The office is then left trying to piece together what actually happened. Why WhatsAppPeople use the easiest communication method available. On construction projects, that is often WhatsApp.

One UK industry article noted that WhatsApp is likely the second most used tool on construction projects after Microsoft Office.It is fast and simple. But it leaves critical information scattered across chats. We have all experienced the frustration of searching for a key message or image. Where AI Becomes Practical AI should not force new behaviour. It should support existing behaviour.

Scopey Onsite uses AI agents, not just generative AI.Generative AI creates content.
AI agents complete tasks and drive outcomes.When someone sends an update through WhatsApp:
- Scopey interprets it
- Understands the intent
- Prompts for missing critical details
- Structures the information properly

We will extend this to email, another place where key information lives.

How It Works
In PracticeScopey turns everyday communication into organised, timestamped, categorised records:
- Disruptions
- Variations
- Quality checks
- Commercially relevant events

For supervisors, nothing changes. They send messages or voice notes as normal. Scopey steps in only when essential details are missing, asking for clarification or a photo. There are no new forms and no new workflow to learn. For the Commercial Team On the office side, events align with the contract.

Commercial teams can see:
- What happened
- When it happened
- What it connects to

Without scrolling through hundreds of messages. We are working closely with our adviser, Mark Kehoe, whose background is in claims and dispute resolution, to ensure the information captured maps to what matters commercially.

The Bigger Picture.
This is not about billing every small change.It is about: Visibility. Trust. Protecting margin. Showing clients the value delivered Resolving issues earlier In a relationship driven industry, clear records support both sides. Better documentation leads to faster resolution.

What’s Next
‍We are building Scopey Onsite alongside design partners in the industry to ensure it is not only valuable, but highly usable.We are launching next year. There may be space for one more design partner, and we will open staged access through a waiting list. If you would like to learn more or see a demo, feel free to connect with me.