Said Achraf

Project: Migration of the B2C Data Platform to Databricks Unity Catalog in an AWS environment. • Supported Feature Teams in migrating data from the Common Data Hub (CDH) to Unity Catalog • Led the progressive migration of data stored on S3 to Unity Catalog using External tables • Implemented the Medallion Architecture (Bronze, Silver, Gold) on Databricks • Created catalogs, schemas, volumes, and governance assets within Unity Catalog • Configured Storage Credentials and External Locations using dedicated IAM roles • Facilitated technical workshops with Feature Teams to structure data and define migration strategies • Defined user groups and managed access control based on business roles • Managed permissions and secured access to catalogs, schemas, and tables • Assisted teams in migrating jobs to Unity Catalog (read/write operations through managed tables) • Configured Unity Catalog-compatible Databricks clusters • Implemented ingestion pipelines from external sources such as Salesforce • Created Service Principals for secure job and pipeline execution • Migrated External tables to Managed tables • Developed Databricks applications to simplify data management and business user data entry • Implemented technical monitoring jobs to supervise workloads and resource consumption • Developed Databricks skills using Genie Code to industrialize the migration of jobs to Serverless mode • Built monitoring dashboards to track assets and detect configuration drifts • Contributed to improving data governance, traceability, and access security

Project: Design and implementation of infrastructure components required to support data use cases in a cloud environment. • Built and maintained the infrastructure of the data platform. • Provisioned Databricks and AWS resources using Terraform. • Implemented CI/CD pipelines on GitLab for infrastructure deployment. • Managed user access and permissions through Databricks groups. • Configured IAM roles to secure resource access and manage permissions. • Built and pushed Docker images to ECR and managed their deployment on EKS. • Ensured operational maintenance and continuous improvement of the infrastructure. • Migrated data from On-Prem environments to the S3 landing zone. • Implemented data ingestion pipelines from the landing zone to Databricks data layers (Bronze, Silver, Gold). • Monitored ingestion jobs through DataDog dashboards. • Resolved infrastructure and pipeline incidents. • Maintained and supported the On-Prem platform environment. • Handled operational requests related to user management and HDFS space provisioning.

Project: Migration of the DPAE project (Pre-Employment Declaration) to a Big Data platform. • Developed Spark (Python) jobs for missing value imputation based on contract type, contract duration, gender, and age • Developed Spark (Python) jobs to generate aggregated reporting tables for pre-employment declarations • Implemented Spark (Python) jobs for duplicate detection, delay rate calculation, and late declaration correction • Developed Sqoop import jobs to ingest data from Oracle databases • Stored data in ORC format on HDFS • Developed HQL scripts for Hive tables using ORC and Parquet formats • Monitored job executions and analyzed logs using YARN Resource Manager • Performed unit and integration testing • Scheduled and orchestrated workflows using Oozie • Provided support to SAS users across the Big Data environment

Project: Automated anomaly detection in trading operations and liquidity ratio correction. • Extracted data from business applications and standardized JSON schemas • Loaded data into the Data Lake and optimized ingestion processes • Implemented an exploratory engine based on 13 unsupervised learning algorithms to detect new anomaly patterns not previously identified by business teams • Performed data augmentation and enriched datasets with engineered features • Developed post-processing modules for outlier handling and anomaly scoring • Conducted statistical analysis and identified anomaly patterns within large-scale datasets • Collaborated with business teams to validate detected anomalies • Deployed scalable machine learning algorithms into production on the Data Lake • Implemented an online learning module for continuous model retraining