Insurance & Agritech

Yield of agricultural fields

The climate subsidiary of a French insurance leader needed to predict field yields across Germany to price a new drought insurance offer. Internal data was not enough. We built the predictive models that made the product marketable.

Problem

The client launched a parametric insurance offer dedicated to European farmers, indexed to drought conditions. To price it correctly, he had to be able to predict field yields by region and by type of crop based on soil moisture levels.

Problem: internal data was insufficient, the necessary climate and geographic variables were scattered across dozens of heterogeneous Open Data sources (Copernicus, ERA5, rasters, shapefiles), and no existing model allowed this data to be cross-referenced to produce reliable predictions at the scale of an entire country.

Vue rapprochée d’une coupe transversale colorée d’une géode montrant des couches concentriques de minéraux en jaune, marron, rouge et vert.

Solution

What we built

We deployed a team of 2 Data Scientists and 1 Lead Data Scientist to design a complete predictive system on the German perimeter.

Step 1 — Multi-source collection and enrichment. Aggregation of internal data (customer business partners) with multiple Open Data sources: satellite data (rasters), geographic boundaries (shapefiles), historical meteorological variables (ERA5, Copernicus). Massive preparation and transformation work to make this heterogeneous data usable together.

Step 2 — Predictive modeling. Two problems treated in parallel: yield prediction by region and by crop (regression), and detection of drought years by region (classification). Benchmark multiple approaches — neural networks, LSTM, LSTM, Random Forest, Gradient Boosting, LightGBM — to identify the most efficient model for each problem.

Step 3 — Selection and deployment. The Random Forest was selected for its optimal performance. The model was deployed via REST API with a visualization layer in the form of geographic heatmaps, allowing business teams to navigate performance predictions by area.

Projects in the same category

See all projects

Supply Chain Optimization Application

A pharmaceutical distribution player had to rethink its entire supply chain: pharmacy assortment, inventory management and delivery channels. Operational research alone was no longer enough. We built the application that transformed its operations.

Social media trend detection

In a market where consumer behaviors change faster than decision cycles, a customer needed to anticipate trends instead of experiencing them. We built the platform that turns social media noise into actionable signals.

Yield of agricultural fields

Customer needs analysis platform

A large French group needed to understand the current and future needs of its customers by aggregating a massive volume of consumer data. After a POC of 8 weeks, we industrialized a complete platform deployed on all French entities and is now being extended at the group level.

Skin scoring and analysis

A major player in luxury cosmetics had developed a skin scoring algorithm. Problem: no one could verify what the AI was based on to make its diagnosis. We built the visualization system that makes predictions transparent and deployable in stores and on mobile.

Beaconing detection

A large group's security team needed to identify beaconing signals, regular network traffic sent by potentially compromised machines to servers controlled by attackers. The volume of logs made manual detection impossible. We built the anomaly detection system capable of processing this data on a large scale.