Oussama Hmidene

• Built and maintained automated data pipelines to ingest livestock, trade, and invoice datafrom40+ international sources using
Python, Selenium, and BeautifulSoup.
• Designed and orchestrated ETL workflows with Dagster, integrating Snowflake and PostgreSQL to deliver clean, reliable, and
analytics-ready datasets.
• Implemented data quality pipelines (freshness, completeness, gap detection, anomaly detection) to ensure dataset accuracy and
consistency.
• Containerized extraction and processing services using Docker, enabling scalable and reproducible deployments.
• Modeled and prepared analytics-ready tables consumed by dynamic Power BI dashboards, enabling insights on production, slaughter
statistics, and tradeflows across countries.
• Supported downstream BI and reporting use cases by optimizing data models, refresh performance, and data availability.
Tech Stack: Python, Dagster, Snowflake, PostgreSQL, Docker, Selenium, BeautifulSoup, Pandas, Power BI, FastAPI

Livestock and Meat Project :
• Developed web scraping pipelines (Selenium + BeautifulSoup) to collect livestock data from 10+ country-level
sources.
• Automated ETL workflows using Dagster and integrated with Snowflake, ensuring timely and clean data
delivery.
• Designed dynamic Power BI dashboards to provide insights on meat production, animal slaughter
statistics, and average carcass weights across different countries.

Trade Agriculture Project:
• Built a Selenium-based web scraping pipeline to extract international trade data from 30+ countries, using a
custom ML model to solve CAPTCHA challenges and containerized the solution with Docker for easy
deployment.
• Cleaned and transformed raw trade datasets using Python, and developed time-series forecasting models to
predict import/export quantities of agricultural products based on historical trends and seasonality.
• Designed dynamic Power BI dashboards tailored for trader clients, providing insights of global trade flows,
price trends, and cross-country comparisons

Invoice Extraction with LLMs and RAG Project:
• Built a system using AI models and few-shot learning to extract key details from scanned invoices of various
formats.
• Used FAISS and prompt design to help the AI handle different invoice layouts without extra training.
• Deployed the solution with Python, FastAPI, and Docker, reducing manual data entry by over 80%.

Data Quality Project :
• Implemented data quality pipelines using Dagster to ensure livestock dataset accuracy and completeness.
• Developed automated checks for data updates, gap detection, and anomaly identification to maintain data
integrity.
• Ensured data met client expectations by validating consistency, accuracy, and completeness.

Technologies: Python, Dagster(Orchestration), Docker, Snowflake, PostgreSQL, Power BI, Pandas,
BeautifulSoup, Selenium, Computer Vision

AI web agent project :

• Developed AI-driven automation solutions for web browsing and repetitive online tasks, such as flight and
public transportation ticket booking.Demo Link showcasing the automation in action.

• Designed a data extraction pipeline using Selenium, integrating Pandas and SpaCy for text preprocessing
and feature engineering.

• Built machine learning models to improve web element selection using LazyPredict and Keras, optimizing
for accuracy and efficiency.

Technologies: Python, Selenium, Pandas, Keras, Scikit-learn, LazyPredict, Docker, Bash

• Built automated web data extraction pipelines using Python and Seleniumto support AI-driven web agents.
• Designed data preprocessing andfeature engineering workflows using Pandas and NLP techniques, preparing structured datasets
fromraw web content. (Demo)
• Developed and evaluated machine learning models to improve web element detection and automation reliability.
• Containerized automation and ML workflows using Docker and Bash to ensure reproducible execution.
Tech Stack: Python, Selenium, Pandas, Scikit-learn, Keras, Docker, Bash

• Led a team of 4 to encourage students to join the AIESEC exchange program and supported them throughout
their experience.
• Worked with the business development team to build partnerships that support AIESEC.
• Created dashboards to analyze student data and developed a ranking system to identify students with the
best chances of being accepted for internships

• Built automated data extraction and pipelinesfromthe Jira API, enabling scheduled updates of productivity and performance data.
• Modeled and transformed Jira data using Power Query and DAX to compute KPIs, performance evolution, and employee rankings.
• Developed dynamic Power BI dashboards to provide stakeholders with clear, interactive insights into teamproductivity and trends.
Tech Stack: Power BI, Jira API, Power Query, DAX, SQL