Getting back more than 80 hours a year on site reports
Site reports made on-site, with speech-to-text and photos, saving more than 80 hours a year.
- Services
- Custom Software
- Platform
- Next.js · TypeScript · Microsoft Azure · Sentry
- Timeline
- 2024

Project Overview
Who is Renoffice?
Renoffice is a Belgian renovation company with more than 35 years of experience. Based in Kortrijk, they work with a team of ten. Every year, they manage more than a hundred renovation projects. Transparent communication, professional project follow-up, and respect for the client's budget are their trademark.
At Renoffice, writing site reports was still done by hand. A coordinator would visit the site, take notes and photos, and write up the report back at the office afterward. With a hundred projects a year, that adds up fast. And every site manager did it slightly differently.
The challenge
Renoffice came to Flowkify without a defined project in mind. Their question was broader than that:
"How do we keep up with A.I.? How do we get real value from it? Are we missing the boat?"
That's an honest place to start. And that's exactly why we didn't begin with building — we began with choosing.
Project Delivery
The approach: choose first, then build
We started with an A.I. Opportunity Mapping. Together with management, we mapped out every idea and plotted them on two axes: value and complexity. Four ideas rose to the top: RenoField, BuildConverse, SupervisorAI, and a Leaderboard.
RenoField came out on top: high value, limited complexity. A genuine quick win.
Next, we built a business case. Not on gut feeling, but on numbers: how many coordinators, how many site visits a year, how much time per report, and what a tablet with a subscription costs. That way, Renoffice knew what the project would deliver before a single line of code was written.
What we built: RenoField
RenoField is an app that lets a coordinator write the site report right there on site — on an Android tablet, using the device's camera and microphone. The report is done before they're back in the car.
1 · Creating the report on-site
- The coordinator opens a new report and selects the project.
- They mark who was present.
- They take photos of the general situation.
- For each part of the site (demolition, structural work, roof carpentry, exterior joinery), they add notes.
The benefit: everything gets captured the moment the coordinator is standing there, not hours later at the office.
2 · Speech-to-text with A.I. polish
- The coordinator speaks their note out loud.
- A.I. converts speech to text.
- A.I. rewrites that text into a clean, professional wording.
The benefit: you don't type on a construction site. Speaking is faster, and the end result reads just as professional every time.
3 · Photos with annotations
- Photos are uploaded straight into the report, organized by section.
- The coordinator marks directly on the photo what it's about.
The benefit: no more back-and-forth about which detail was meant.
4 · Messaging module for subcontractors
- The coordinator speaks what they want to say.
- A.I. turns it into a proper message, addressed to the right subcontractor.
- They check it and send it.
The benefit: follow-up with subcontractors happens straight from the report, not in a separate chat.
5 · Report as a PDF
The full report is automatically turned into a PDF, ready to send. All reports are kept together in one overview.
The tech
RenoField is a full web app, not an off-the-shelf package. We built it on Next.js with TypeScript, ShadCN/UI, and Tailwind CSS, with NextAuth for login and Sentry for monitoring. Hosting and the database run on Microsoft Azure. The app works on an Android tablet as the primary device, and in the browser on a laptop as a second channel.
We deliberately chose to build a custom app. Off-the-shelf tools for site reports exist, but none of them combine voice input, annotation, and direct messaging to subcontractors.
The timeline
The exploratory A.I. advisory track started in April 2024. The business case followed in September 2024, and the MVP build kicked off at the end of September 2024. From idea to working app took less than eight weeks. During the build, we worked with an open backlog and a weekly report of risks and actions, so Renoffice always knew exactly where the project stood.
Results
The business case projected more than 80 hours saved a year. That comes from a simple calculation: four coordinators, forty work weeks, half an hour less work per site visit. For a company of ten people, that's roughly two full work weeks freed up.
Beyond the time saved, there's the quality. Site reports now look consistent everywhere, with photos and annotations in the right place. What gets said on site actually ends up in the report. Employees, clients, and subcontractors all notice the difference.
RenoField is one of the first applications Flowkify ever built, and it still runs in production every single day. Flowkify continues to support Renoffice as its A.I. partner, managing the infrastructure and leaving room to keep building on the app.
Want to find out what A.I. could do for your site follow-up? Get in touch with Flowkify for a tailored A.I. scan.
Ready to build your case?
Tell us where it's clunky. We'll design the flow and build it, often live within a few weeks.
No-obligation call · reply within 24 hours
Other cases
More work that delivered results.
Dewulf GroupTechnical3 zonesfor A.I. and self-build
How Dewulf Group is getting a grip on A.I. and Copilot with a governance projectA.I. governance and Power Platform governance for an international machine builder.View cases MerakiAccounting50%time saved per tax advice
How Meraki went from A.I. trainings to a prioritized A.I. roadmapFrom A.I. training to a prioritized A.I. roadmap for an accounting and advisory firm.View cases ExplicietMarketing70-75%of note-taking work handled by A.I.
How Expliciet automates meeting notes with an A.I. notetakerAutomatic meeting reports in SharePoint for a marketing agency with 40 employees.View cases
