B SMART: the mayor of Le Port-Marly and HelloSpot on AI for residents’ requests
On Smart Impact (B SMART), Cédric Pemba-Marine, mayor of Le Port-Marly, and Dominique Péré, CEO of HelloSpot, explain how AI helps a town hall handle residents’ requests.
On November 9, 2026, Smart Impact, a programme on the French business channel B SMART, asked a simple question: how can a town hall better follow its residents’ needs? Host Thomas Hugues welcomed Cédric Pemba-Marine, mayor of Le Port-Marly (Yvelines), and Dominique Péré, CEO of HelloSpot, the AI-powered citizen relationship management platform.
Watch the programme on B SMART (13 minutes, in French)
This is HelloSpot’s second appearance on the channel, after the Smart Tech interview in February on responsible AI.
In Le Port-Marly, it all started with an emergency
Le Port-Marly, a town of nearly 6,000 residents, worked like many municipalities: with paper and a spreadsheet. Then came two events officially recognised as natural disasters, including a landslide. Around 200 people had to leave their homes, and the town hall found itself with no tool to follow their rehousing requests.
When Dominique Péré asked which issue caused the most difficulty, the answer was immediate: rehousing the people affected. No software on the market is built for that. This is where AI makes the difference: it lets the tool adapt to the town’s specific case. “A town hall is a world of its own. And a different world every time,” he sums up.
Requests coming from everywhere
The mayor sees it every day: residents speak up more than before, and expect a lot from their town hall. “Very demanding. And they have every right to be. That’s what we’re here for.”
These requests come through the switchboard, through emails sent to different addresses, through the town’s app, or simply in the street, when a resident stops an elected official. “There are clearly 10,000 ways in for information,” notes Cédric Pemba-Marine. The real challenge is on the way out: finding the right person to handle each request.
That is the role of citizen relationship management: every request arrives in the same place, whatever the channel, and each one is assigned and tracked until it is answered.
Bringing weak signals to the surface
HelloSpot’s AI interprets each request as it arrives, assigns it to the right staff member and suggests a reply. It also brings weak signals to the surface: requests that are few but important, repeated requests left without follow-up, or another problem hidden behind a resident’s request.
For the mayor, these signals tend to confirm his instinct as an elected official, while revealing issues he would not have seen alone. Dominique Péré points to the hardest request for a town hall to handle: the complaint that follows no set process. Following it better means entering a loop of continuous improvement. HelloSpot’s statistics then give elected officials concrete material for their decisions.
AI that suggests, staff who decide
Thomas Hugues raised the question of AI errors. Dominique Péré answered on three points:
- French AI: HelloSpot runs on technology from Mistral, within a framework compliant with the GDPR and the AI Act.
- The right context: the AI works on the case at hand and the regulations that apply to it, no more, no less.
- The final word to staff: “AI doesn’t have the final word. Staff check the case.”
On time saved, Dominique Péré estimates that AI halves processing time on average, and that the more complex the case, the bigger the gain. For Cédric Pemba-Marine, what matters most is elsewhere: freeing up time for the human relationship, which AI will not replace. HelloSpot’s AI assistant is designed in that spirit.
For small municipalities too
HelloSpot is designed for local municipalities, from a few hundred to several tens of thousands of residents. In a small town, the need for a tool is the same: the less you do by hand, the more time you have to talk to residents. More than 700 town halls already use HelloSpot.
The free version of HelloSpot lets you get started with up to 1,000 requests, with no public tender and no spending commitment.