Victor Nitu

The Good Apparel is an ethical fashion marketplace. They group a wide range of brands all over Europe. The biggest challenge they face is to aggregate all the different products from different suppliers while keeping a homogenous and coherent online shop. As a data engineer, I had to build a data pipeline that automatically aggregates all the products from various suppliers. I made an administration interface to allow the operational team to manage the products' descriptions and tags. With the help of a machine learning engineer, we plugged the administration interface to AI services to extract tags from products' pictures and to generate product descriptions.

Millenium is the new ERP developed by Foncia Stark for the Foncia Groupe replacing a 15-year-old ERP system. Foncia Groupe is one of the leading real estate company in France. With approximately 550 agencies, it manages about 45,000 buildings and 365,000 rental properties. As a lead software engineer, I managed a team in charge of building the accounting module that would handle the general accounting of Foncia Groupe. With my team, we developed the APIs that would be called by other modules to automate the accounting process. We also developed a specialized administration interface for the accountants of Foncia Groupe.

Figaro Classifieds recruitment department owns different recruitment applications such as Viadeo, Cadremploi, Keljob, and more. Each one of these products contains a section with companies profile pages to present and advertise their enterprise. As a lead software engineer, I developed the internal application to aggregate the enterprises' information and expose APIs to consume this data by the different recruitment products. I also built a centralized administration interface to manage and update each enterprise's identity.

Phi Suite is an event-driven data engineering platform. This platform offers a low-code solution to build event-sourced applications to collect and preprocess data. Optionally, it provides features to connect the generated data to machine learning models and to trigger the learning process automatically.

This open-source project is an example showing how to build an application using PHI architecture. This architecture is a combination of event-sourcing, and CQRS abstracted into three core concepts: Inspector, Handler, and Publisher. It was build in Python using Kafka and MongoDB. The medium article "Event-Driven Data Collection" explains every detail of this project.

This open-source project showcases how to use event-sourcing to obtain real-time data visualization. It contains a dashboard build with Svelte and Chart.js and a Jupyter notebook using Matplotlib. The medium article "Event-Driven Data Visualization" explains in detail how to achieve those results.

This open-source project shows how to extract data from MongoDB and Kafka and preprocess it to obtain the required format. This process is a common use-case when working with artificial intelligence to transform the data into the correct format to train machine learning models. You can find more details in the medium article "Event-Driven Data Preprocessing."

This open-source project is a lightweight library to build gRPC APIs in Go. It handles all the standard configuration and automatically includes a health check functionality. It makes it easy to start a gRPC server quickly.

This open-source project is a small and lightweight GraphQL client library. It was built in Typescript to be usable in a Typescript and JavaScript project. It only uses the native fetch API under the hood.

This open-source project is a lightweight client library to communicate with a Neo4j database. It handles the complexity of creating transactional Cypher queries. It is built on top of the official Neo4j driver for Go.