Aksel Ayoub B.

I design and industrialize scalable data pipelines and backend systems using Scala, Python, Spark, and Hadoop.

My work focuses on building high-performance, distributed solutions for large datasets, ensuring code is clean, reliable, and production-ready.

Core skills:

Big Data processing with Spark & Hadoop

Backend development in Scala & Python

Workflow automation and data pipelines

System optimization for scalability and performance

I bring a strong engineering mindset, bridging complex models and real-world, maintainable software.

Project Overview:

After Amplitude Studios' transition from SEGA, Code Busters was engaged to create a robust AWS and Databricks data platform designed to handle billions of HTTP requests monthly.

Role and Responsibilities:

As a cloud data architect, my key accomplishments include:

1. Proposal and Architecture Design:

Helped developing a comprehensive RFP response outlining a scalable AWS and Databricks architecture tailored to Amplitude's needs, including a cost analysis.

2. Cloud Infrastructure Implementation (AWS):

Infrastructure as Code (IaC) with Terraform:

Established a public endpoint using AWS Application Load Balancer (ALB).

Developed and deployed an API for capturing HTTP requests via AWS ECR and Fargate.

Implemented AWS Kinesis Firehose for streaming and buffering.

Configured AWS S3 for JSON data storage.

Managed network security through AWS VPC and defined user permissions with AWS IAM.

Monitored and optimized costs.

3. Databricks Infrastructure Setup (IaC with Terraform):

Configured Databricks, including storage, credentials, user groups, and clusters.

Designed data ingestion pipelines (raw-to-bronze, bronze-to-silver, silver-to-gold) using Databricks Spark jobs and Delta Live Tables (DLT).

Implemented Unity Catalog for enhanced data governance.

4. Historical Data Migration:

Shared historical datasets using Delta Sharing within Unity Catalog and enhanced existing jobs for historical data ingestion.

5. Load Testing and Monitoring:

Analyzed traffic patterns and simulated loads with Locust.

Adjusted Fargate autoscaling for optimal API request handling.

Optimized Firehose buffering for manageable Spark files and created monitoring tables for job execution statuses.

Tech Stack:

Cloud: AWS (ALB, ECR, Fargate, Kinesis, S3, VPC, IAM)

Data Processing: Databricks, Spark, Delta Live Tables (DLT), Unity Catalog

Infrastructure as Code: Terraform

Load Testing: Locust

Data Engineer at Engie Global Markets, specializing in forecasting models for residential and business energy consumption (power and gas). I collaborate with quants and data scientists, leveraging ETL pipelines managed by Airflow and Databricks jobs coded in PySpark.

My focus areas include:

• ETL development (file/API ingestion, complex Scala transformations)

• Collecting various types of data (marketing data, readings, load curves, etc.)

• MLOps and automation

• Monitoring processes and serving ML models

• CI/CD deployment and maintenance

• Performance optimization and testing

• Used Terraform for deploying resources on AWS.

Technical Stack:

• Python (Spark, etc)

• AWS (S3, EC2, IAM, Athena, EMR, Glue, etc)

• Databricks

• Azure DevOps