Tarek Samaali

Pillars: Data Trust, Market Insight, Performance, Optimization, Prediction.

Working on different Proofs of concepts related to computer vision, predictive analysis, NLP and Sentimental Analysis. Project 1: Implemented a flight delay prediction tool for an airline company. Project 2: Built a model for detecting themes and sentiments on textual data specific to a car brand. Project 3: Participated in the development of a preliminary proof of concept for multimodal transport services. Project 4: Implemented Key performance indicators within a Dashboarding tool for a pharmaceutical distributor. Project 5: Took part in implementing a Similarity Search Platform and using it to perform facial recognition and automative Question Answering.

- Used Detectron2 (developed by Facebook) to detect infringing cars in speed camera pictures. Key Words: Python, Image Processing, Computer Vision, Detectron2. - Provided additional ML-based functionalities to an OCR-based solution : Key Words: Python, NLP, Computer Vision ( YOLO ).

- Wrapping a speech-to-text solution into an API web framework and industrializing it on Azure using GPU resources. Key Words : Docker, FastAPI, Azure.

IOT project aiming to collect data from different sensors attached to home doors and building intrusion detection models.

- Provided a data model to leverage different levels of geographical-based analysis. - Compiled statistics related to different sales activities into a Power BI dashboard template. Key Words : Power BI, Power Query, DAX.

- Implemented additional Sales and performance indicators into a dashboard Power BI template. - Attempted to plug the dashboard into a unified Databricks Platform. Key Words : Python, Power BI, Azure Storage, Databricks.

- Developed data-gathering and pipelining routines to test a price- predicting model which is already put in production. - Analyzed its performance and drew up recommendations with regards to its future use and potentiel improvements. Key Words : Python, Snowflake, SQL, DataBricks.

Built a preliminary proof of concept to predict and alert on Sneakers bestsellers using Sales data. Key Words : Python, Machine Learning.

- Provided comprehensive correlation analysis of train routes to address the overconsumption issue of some components of the Metro Line 1 rolling stock. Key Words : Python, Machine Learning, Correlation & Regression Analysis.

- Investigated available resources to develop more useful project plans regarding intelligent fraud detection within suspicious credit transfers. - Took part in different brainstorming workshops to build necessary variables upon which ML models were built. - Implemented a micro-service brick with a view to optimizing a part of the collection, processing and scoring architecture. Key Words : Hadoop, Spark, Hbase, Python ( H2O ), Scala, Machine Learning.

- Performed a benchmark for Textual classification. - Attempted basic summarization of items contained within textual descriptions collected from e-commerce web sites.

Writing a Maven-based code-generator plugin using annotations. Key Words : Java, XML binding ( DOM4j, JAXB ).