Back-end internship · Credit Agricole

Credit
Agricole.

A back-end internship at Credit Agricole: automating data processing to anticipate natural risks based on geographic location.

Brief at a glance
PeriodAug – Sept 2024
ArrangementBack-end internship
TypeInternal tool · data analysis
StackOOP back-end · API
StatusInternship completed
Screenshot · credit agricole analysis tool
One number, that’s it

Processing time, cut down.

Several hoursa few seconds

Manual processing time for a data file, brought down to an automated calculation.

The context

Hundreds of addresses, 9 risks to check.

Before every property financing decision, Credit Agricole teams had to assess natural-risk exposure for each property: flooding, earthquakes, ground movement, clay shrink-swell, forest fires, radon, hazardous pipelines, soil pollution, and dam failure.

The process was entirely manual: look up each address, download the official Géorisques report, read through the PDF page by page, copy the data into a spreadsheet. For a portfolio of several hundred addresses, this took hours.

What was built

A file goes in, a full report comes out.

The tool takes a simple list of addresses as input. For each one, it automatically locates the property on the map of France, retrieves the official natural-risk report, extracts the relevant data, and compiles everything into a ready-to-use table.

Nine risk categories are analyzed per address, with two levels of detail: risk at the exact location and risk at the municipality level. Results are then displayed on an interactive map for an at-a-glance overview of risk zones.

The outcome

Hours of work cut down to seconds.

What used to take a full day for an address portfolio now runs in the time it takes to grab a coffee. Teams get a structured table with all 9 risks per property, ready to plug into their decision-making tools.

The tool freed advisors to focus on analysis and client relationships rather than data collection, and ensured no risk gets overlooked for lack of time.

Tech stack.

Language
Python
Data
Géorisques API · OCR
Geocoding
Adresse API (gouv.fr)
Output
CSV + interactive map

A data or
back-end project?

If you need to automate data processing or build a reliable API, let’s talk.

Direct booking
FormatDiscovery call
Duration30 minutes
Response< 24h
CommitmentNone
Book my slot →
[email protected]Toulouse, FR
© 2026 Noah Vernhet. All rights reserved.Designed & built in Toulouse