airbus

Context Airbus manufacturing teams needed to track KPIs such as production output, defect rates, and workload distribution across multiple internal data sources. Reports were generated manually each month using spreadsheets and required significant effort from analysts.Problem The lack of a centralized, automated system resulted in duplicated work, inconsistent metrics, and slow access to critical operational insights. Teams had limited visibility into trends and comparisons across production lines and sites.Solution I developed an end-to-end analytics pipeline to automate the collection, modeling, and visualization of manufacturing KPIs:<ul><li>Airflow orchestrated the extraction and loading of raw operational data into Skywise</li><li>dbt was used to clean, model, and test data in a modular and scalable structure</li><li>Aggregated KPIs were used to feed existing Skywise dashboards, avoiding the need to rebuild new reports monthly</li><li>Filters and drilldowns allowed teams to explore trends by site, product line, and time period</li><li>The pipeline was designed to accommodate evolving data structures based on subcontractor input</li></ul>Results<ul><li>📉 Significant time saved on monthly reporting</li><li>📊 Standardized KPIs across teams and sites</li><li>⚙️ Improved reusability and dashboard sustainability</li><li>🧱 Robust, scalable pipeline supporting operational decision-making</li></ul>Stack Airflow, dbt, Skywise, SQL, Python, Git

Data Engineering

DevOps

 Automated KPI Monitoring Pipeline for Manufacturing Operations (Airbus)

Airbus

Context Design review processes at Airbus involve frequent checks on technical modifications, many of which turn out to be unnecessary. These reviews are time-consuming and resource-intensive, especially when working with varying data from external subcontractors.Problem The goal was to develop a tool to identify which design changes truly require review — and which could safely be skipped — in order to save time and reduce operational costs. A scalable, reliable pipeline was also needed to handle evolving data structures from multiple subcontractors.Solution I first designed a robust, modular pipeline using dbt, ensuring data from changing sources could be cleaned, standardized, and versioned automatically. Once the data foundation was in place, I trained and validated an XGBoost model to predict whether a design change should be reviewed. The tool was deployed as a user-facing application integrated into the team's workflow and is now used daily by engineers.Results<ul><li>⚙️ Reduction of unnecessary design checks</li><li>💰 Time and cost savings for engineering teams</li><li>🧱 Scalable and maintainable pipeline ready for changing data sources</li><li>✅ Model fully integrated into day-to-day operations</li></ul>Stack dbt, SQL, XGBoost, Python, Git, internal deployment tools

Talel Taieb

Data Analyst & Data Engineer | dbt • Airflow

About me

Work experience

Data Analyst

Education

ENSEEIHT

INSA Toulouse - Institut National des Sciences Appliquées de Toulouse

Portfolio

Automated KPI Monitoring Pipeline for Manufacturing Operations (Airbus)

airbus

ML-Based Decision Filter for Design Review (Airbus)

Airbus

Talents

Recruiters

About