Narjes Atia

-Built robust and scalable data pipelines to automate ingestion, transformation, and consolidation of data from SQL databases, flat files, and APIs — reducing data availability time for business teams by 40%. -Built optimized storage layers in Azure Data Lake Storage Gen2 using Delta Lake format to support ACID transactions . -Implemented comprehensive data quality checks (uniqueness, format validation, cross-field consistency), enabling early detection of 95% of data anomalies . -Secured pipelines and resources with Key Vault-based secret management, managed identities, and role-based access control (RBAC) to ensure compliance and data protection. -Built custom monitoring dashboards with Azure Monitor and Log Analytics, enabling real-time visibility into pipeline performance and SLA tracking -Collaborated closely with BI teams to provision clean, governed, and well-modeled datasets for Power BI reporting and decision-making -Streamlined deployment processes using CI/CD workflows, reducing release times and minimizing regression errors in production.

-Led the design and implementation of end-to-end ETL pipelines to ingest, clean, and consolidate high-volume citizen data from diverse systems including civil registries, social services, and national identification platforms. -Played a key role in the RIUC project (Registre d’Identifiant Unique du Citoyen), a strategic public-sector initiative aimed at assigning a unique ID to every citizen. This involved consolidating fragmented data across ministries, aligning inconsistent schemas, and resolving identity conflicts using probabilistic matching and rule-based logic . -Designed PostgreSQL data models for storing master data, surrogate keys, and reference mappings. -Developed automated data quality rules to flag duplicates, detect structural inconsistencies, and validate data integrity before integration. -Collaborated with legal and compliance officers to implement data privacy and access control mechanisms, ensuring that data sharing across entities met national data protection regulations. -Delivered technical documentation, data dictionaries, and onboarding material to facilitate system handover . -Led a small team of engineers and data scientists, assigning tasks, reviewing code, and validating models to ensure high delivery standards.

-Designed and implemented advanced AI models to analyze massive collections of judicial documents (court rulings, legal notices, case files), enabling semantic understanding and automated classification at scale. -Designed and developped NLP techniques such as Named Entity Recognition (NER), document embeddings (BERT, TF-IDF), and semantic similarity scoring to extract legal entities and identify related or duplicate cases. -Optimized inference workflows for both batch and on-demand processing, reducing API response times by up to 60%. -Collaborated closely with legal experts and business stakeholders to translate domain knowledge into actionable data features and model outputs.

Development of a cross-platform application Airports Tunisia , a useful application for the realtime flight status in Tunisia ( arrival and departure) Exploring FlightStats web services APIs .