Leadership:
Leading a team of 5 Technical Leads, who collectively manage a group of 20 software engineers, ensuring project delivery, cross-team collaboration, and alignment with organizational goals.
Initiatives:
- Introduced and standardized the usage of Private AWS API Gateway across all Valeo digital tool chain
- Designed the first hybrid system in Valeo which combines GCP and AWS to help Valeo R&D team to work closely with BMW and other OEMs to co-engineer electric cars software
- Introduced AWS Step Functions to handle a complex workflow solution
Achievements Highlights :
- Currently leading the migration of a monolithic application to microservices architecture.
- Design and implemented AWS Private API Gateway to improve API Management for private APIs hosted in Valeo VPCs, the private API gateway connects to private APIs in Valeo through AWS VPC Link and is only accessible from Valeo hybrid network through Interface Endpoints so that traffic doesn't go through the internet
- Designed data pipelines which are responsible for extracting data from different data sources ,load them into Valeo data warehouse hosted in BigQuery and transform them to be ready for modeling and KPIs calculations, the orchestration of this data pipelines is done through Tibco Business Works and the data extraction is done through dedicated REST APIs on top of each data source
- Designed a hybrid GCP + AWS + On-Premise architecture for a package ingestion pipeline that allows Valeo to securely receive terabytes of packages through Valeo's standard APIGEE Gateway from external partners to Valeo's private AWS VPC without exposing it to public access
In GCP , APIGEE Gateway is used to securely expose endpoints to the customer using signed urls, with which the customer shall upload the packages to an encrypted Google Cloud Storage bucket
The GCS Bucket is a publisher to a pub/sub topic for which it publishes the file upload event metadata
In AWS Valeo VPC, an always running container hosted on ECS Fargate is subscribed to the GCP Pub/Sub topic to react to the file upload events and starts downloading the packages from GCS, then it triggers an AWS Step Functions Workflow
AWS Step Functions are used as a serverless workflow manager with a poll model, where workers poll for the job status
Other steps in the workflow are a mixture of Java Spring services hosted on ECS Fargate Containers and EC2s
Packages are stored in an HSM S3 bucket
DynamoDB is used to persist the workflow traces