Mobility Data Analysis

During the first year of my master’s program at Aix-Marseille University, I had the privilege of interning at Mobility Metrix, an innovative startup specializing in mobility data analysis and developing business software for data-driven decision-making and performance monitoring. This opportunity marked my first step into the professional world of data science, allowing me to apply my academic knowledge to real-world challenges in the field of mobility.

Working with Heterogeneous Data

One of my primary responsibilities during the internship involved analyzing data from a variety of sources. The data was highly heterogeneous, requiring a meticulous approach to handle its diverse origins and formats effectively. The sources included:

• APIs from third-party applications: These provided structured but inconsistent datasets that required standardization to ensure compatibility and integration with other data sources.
• Open data from transportation networks: These datasets varied significantly in format, granularity, and quality, reflecting the differing standards and methodologies of the organizations providing them.

Each dataset posed unique challenges, including issues of incomplete records, varying data structures, and inconsistencies in measurement units or timestamps. My role involved:

1. Data Cleaning: Identifying and addressing anomalies such as missing values, duplicate entries, and erroneous data points to improve the datasets' integrity.
2. Data Transformation: Standardizing formats, reconciling disparate schemas, and ensuring all datasets were compatible for downstream analysis.
3. Data Quality Assurance: Implementing validation checks to verify the reliability and consistency of the processed data.

This preparation was essential to make the data usable for meaningful analysis and to support data-driven decision-making of the company's clients.

Building the Database

Beyond analysis, I contributed to designing and implementing a PostgreSQL database to store the cleaned and relevant data. This involved:

• Structuring the database schema to accommodate diverse datasets.
• Writing scripts to integrate processed data into the database efficiently.

What I've Learned

Working at Mobility Metrix was not just about technical tasks, it was a crash course in adapting to a fast-paced, startup environment. I learned:

• How to collaborate with a team of professionals.
• Practical approaches to handling messy, real-world data.
• The importance of creating scalable and maintainable solutions.

This internship laid a strong foundation for my journey in data science, providing me with hands-on experience and a clearer understanding of the industry’s demands.