Datalake developpement in 3 months

Context

We assisted a start-up specializing in geo-contextual intelligence in developing a data architecture capable of ingesting a large volume of data.

The Challenge

The company had two different data sources: information on points of interest (POIs), such as stores and museums, as well as mobility data. The first challenge was to provide an aggregated view of these two data sources, thereby managing a significant data volume from the outset.

Next, we needed to develop a solution that could address their clients' concerns, such as:

What is the most popular store?
Are my visits increasing?
Which stores do customers visit before/after visiting mine?

The architecture had to be flexible enough to integrate additional sources, such as weather conditions, to enrich the data further. Furthermore, this start-up had a small team of just two people to maintain the solution.

The Solution

We developed a data lake in the GCP cloud, capable of ingesting various data sources while ensuring their quality, allowing for enrichment, association, and extraction in different file formats. Additionally, we helped the start-up build a dedicated Data team.