Alaeddine Ben Cheikh

- Worked for a major french energy company on building and maintaining a datalake from scratch. • Developed efficient data transformation workflows using python and aws services (s3, lambda, SQS, SNS, StepFunctions, Redshift, ECS, ECR, AThena, Glue, Batch, Fargate, CloudWatch, EventBridge). • Implemented multiple services infrastructure using Terraform. • Successfully scaled existing data pipelines on aws batch and ecs fargate infrastructure. • Designed and implemented a robust data architecture to resolve the data platform scaling issues. • Actively participated in sprint planning sessions. - Preparation and delivery of training content for engineers, including designing the course materials and conducting the training sessions. - Managed and mentored junior consultants and trainees.

- Worked for a major French insurance company on automating the response to the credit limit request made by customers. • Participated in interviews and workshops with business teams to understand workflows, design data warehouse tables and validate data quality. • Developed ETLs to feed a data warehouse used for data science and analytics uses cases. • Designed KPIs and developed analytics solutions to support decision makers. • Built different modelling approaches to automate the credit limit response process. - Worked for a major media company to build a recommender system. • Used aws stack to deploy models in production. • Developed and refactored data pipelines. • Code and ML architecture optimization. - Developed and deployed scoring model for a major French luxury company to identify bag resellers in the secondary market. - Built and deployed a recommender system using GCP stack for American tourists to offer them an original touristic guide in Paris based on their preferences.

• Project : Digital strategy consulting for an European business school & Predictive Lead Scoring - Feature Engineering : building custom features from a CRM database ( cleaning, imputation, coding categorical features, removing outliers and duplicates ... ) - Modeling : building binary and multi-class classifiers in order to predict the best next product for the prospect. -> Techniques : Random Forests, Xgboost -> Tools : Python, Scikit-learn - NLP : Using natural language processing techniques to clean some features and to process and classify linkedin text data. -> Techniques : Keywords detection (tf-idf), topic modeling, Sentiment analysis, n-grams -> Tools : Python, word2vec, FastText

- Getting data, extraction, automating and optimizing the data processing in streaming and batch mode. - Developing an extension to the Appdynamics Tool in order to supervise logs of applications that deliver services of Bouygues Telecom. - Initializing a Poc using ElasticSearch in order to predict anomalies in the production environment. Tools : Java, Python, ElasticSearch, Kibana, AppDynamics. Techniques : Anomaly detection, Web scrapping