Salah - Architecte Cloud AWS

Team Lead & Senior Solutions Architect

Missions: Leading the architecture, migration, and development of Air Liquide’s Data Platform on AWS.
• Leading and collaborating:
- Led the migration of the platform from Air Liquide’s corporate host to Claranet’s AWS host (Because Claranet
holds the HDS (Health Data Hosting) certificate, it was chosen by Air Liquide to deploy the platform in production to be
HDS compliant)
- Worked tightly with the head of the data platform and Claranet’s team to structure the migration of the
platform
- Succeeded in leading the team (+10 pers) to put two use cases into production under a very challenging
deadline
- Led and coordinated meetings with different stakeholders, teams, and product owners to discuss, architect,
and evolve the data platform
- Established backlogs and organized the tasks’ dispatching to different actors of the data platform
• Architecting:
- Challenged and optimized the existing architecture of the platform (cost optimization, resource’s distribution,
API calls, networking, security and IAM roles, etc...) to harden the platform’s operability and pare down major
incidents.
- Produced architecture documents which became the reference to go to when auditing the platform
- Architected and developed new features on the data platform (different connections to on-premisessystems:
data replication with DMS and Qlik Replicate, file transfers to on-premises systems, etc…)
- Designed and developed different CloudFormation stacks of the data platform
- Developed automation tools to accelerate the platform’s deployment (script generation, etc.…)
- Deployed Jenkins pipelines to automate the platform’s deployment on different environments

Data Engineer & Architect

Client: Mutex: Succeeded in building and deploying an AWS data platform from scratch into production in 6 months.
The platform is meant to centralize, process, manage insurance data and offload legacy systems.
Missions: - Contributed to designing the infrastructure layer of the platform: S3 as a DataLake, Redshift as a Data
Warehouse, Glue for cataloging the data, step functions for workflows
- Implemented elements of the platform's infrastructure with Terraform
- Architected the different patterns of the Data platform (batch ingestion, cataloging, etc...)
- Implemented the data cataloging pattern: used Glue’s Classifiers, Crawlers, and jobs for this pattern
- Designed and implemented ingestion and cataloging workflows with Lambda functions and Step Functions
- Designed and implemented the data traceability pattern

Technical Lead & Architect

Client: Accenture: Set up an accelerator to train and deploy Machine Learning models in production.
Missions: - Reviewed the infrastructure requirements (VPCs, Security Groups, Subnets…) and their Terraform
implementations
- Implemented Machine Learning Models with MLFlow
- Designed and implemented the pipeline to deploy Machine Learning Models (from the commit till the Model
deployment on Sagemaker
- Supervised a team of two persons to design and implement the application components

Big Data Engineer & Architect

Client: Crédit Agricole Assurances: Set up of a Big Data platform (MapR) to centralize, process, manage insurance data
and offload legacy systems.
Missions: • Architecting:
- Built a Data Platform for managing insurance data and solving very ambitious use cases
- Designed and supervised 3 use cases in production environment
- Designed and implemented data pipelines with respect to architectural patterns of the Data Platform and
the existing Information System
- Architected an Elasticsearch cluster: proposed an architecture for an on-premise geo-cluster (Two Data
Centers in two different locations) which provides High Availability and Fault Tolerance
- Integrated new technologies and capacities in the Data Platform by studying Conceptual, Logical and Physical
data models, and implementing solutions in Java
• Implementing:
- Identified and solved bugs in several environments
- Solved real time streaming challenges
- Mastered the MapR distribution and the Hadoop stack
- Provided support and knowledge transfer for other teams

R&D Engineer – Built a Complex Event Processing (CEP) Platform

R&D thesis graduation internship (6 months) around Complex Event Processing.
- Defined & dove into the meaning of events and the basic concepts of Complex Event Processing (CEP)
- Analyzed the market vision and benchmarked the existing CEP solutions.
- Designed a functional and technical architecture for a CEP solution.
- Analyzed the correlation between CEP and AI, Big Data & Cloud
- Realized a Proof Of Concept for a specific use case (anomaly/Fraud detection), use of machine Learning Algorithms with
an existing solution.

Student Research Assistant – Built a mini-ChatBot

- Implemented a mini-chatbot which helps answering questions of students wanting to integrate the Data science
master’s degree program proposed by Ecole Centrale.
- Dove into concepts of deep learning & natural language processing
- Studied and cleansed data (a corpus of questions and their answers)
- Designed a performant neural network
- Optimized the model performances
Key words: Neural Networks (Sequence to Sequence Networks), NLP, Pytorch, Python

Internship – Web Development | Data Visualization and Project Management

- Designed and developed new features for a creative Web solution which went into production during the first month
- Designed relational databases and several data visualizations
- Created different types of tests: unit tests, end-to-end tests

- Designed and Developed a Credit Simulator
- Interfaced an existing internal SOAP WebService