Bilel Affes

Within a backend team of 7 people dedicated to user management, I contributed to the evolution of theRTLDEstreaming platform by modernizing backend services, improving their reliability, and supporting the migration of the architecture to AWS.

I participated in themigration of services from PHP to Go,a refactoring that reduced memory consumption by **60%**, CPU usage by **56%**, and latency by **70%**, significantly improving the platform's performance, stability, and operating costs.

In parallel, I developed new features, resolved production incidents, and conducted root cause analyses to enhance service reliability.

I also designed and developed acatch-up / replayservice for reliably replaying business events and reprocessing data, facilitating recovery after incidents and ensuring the consistency of asynchronous processing.

Finally, I contributed to the migration of event streams fromVector to DynamoDB Streamsby adapting data models to the AWS ecosystem, as well as developing new Databricks pipelines and defining a more scalable data architecture.

Stacks:Go (Golang), Rust, AWS, DynamoDB Streams, Databricks, Kubernetes, Terraform.

Within Agile squads of about ten people (backend, frontend, PM/PO), I contributed to the evolution of the Molotov streaming platform, used by over 400,000 paying subscribers and processing up to 1,500 requests per second.

Molotov:Contribution to the reliability and evolution of the streaming platform by ensuring backend service maintenance, resolving critical bugs, and developing new features. I notably participated in the implementation of the live rewind feature (4 hours), offering a smooth playback experience, as well as integrating new Smart TV models to extend the platform's compatibility.

UP (Molotov by Fubo):Participation in unifying the backend platforms of Molotov and Fubo to offer a common architecture for the European and American markets. Development of a central backend service ensuring communication between the frontend and microservices, then co-design of a WebSocket service for instantly streaming content from a smartphone to a Smart TV. I also participated in the deployment and integration of this service on Google Cloud Platform (GCP) to ensure its availability and scalability.

Throughout these projects, I worked in high-traffic distributed environments, focusing on performance, service reliability, code quality, and continuous delivery.

Stacks:Go (Golang), Redis, MariaDB, RabbitMQ, Kubernetes, Terraform, AWS, Google Cloud Platform (GCP), WebSocket.

Within an Agile team of 6 people (3 backend developers, 2 system/infra engineers, and 1 manager), I contributed to several projects for telecom operators, developing backend services, event-driven architectures, and cloud solutions.

In parallel with development, I participated in the administration and deployment of infrastructure on Flexible Engine, automating deployments via Ansible and setting up highly available environments.

Smartlife (IoT):Development of REST microservices and unified APIs for controlling connected objects from different brands via a mobile app or voice assistants. Integration of an API Gateway capable of handling up to 1,500 requests per second and implementation of real-time communication between equipment and backend services.

SecureOn (Cybersecurity & Parental Control):Development of backend services for real-time detection and blocking of malicious URLs, as well as parental control features. Performance optimization using Redis in addition to MongoDB, reducing response time by 70% and database load by 65%.

Energy Optimization Platform:Development of an event-driven architecture to collect, transform, and store data from connected objects. This refactoring reduced response time by 80%, decreased MongoDB load by 70%, and improved request throughput by 10%. Development of APIs used for visualizing and analyzing energy consumption.

Stack:Node.js, Python, Redis, MongoDB, MQTT, React, Ansible, Flexible Engine.