Firas Ayari

Passionate ICT engineering student with a strong foundation in data science and AI, eager to leverage technical skills to drive innovation and solve complex problems. Ready to take on challenges and contribute to impactful projects in a dynamic environment

AI Engineer at Originator (2026-06 – Present)

• Design complex and realistic tasks for the training, evaluation, and improvement of LLM-based and autonomous agent systems.
• Develop and improve evaluation pipelines to assess agent performance across diverse execution environments.
• Build and maintain Python-based verification tools to validate agent outputs and identify failure modes.
• Analyze agent runs, prompts, and workflows to identify performance issues and improve task quality and evaluation reliability.
• Collaborate with technical and quality teams to design, validate, and improve AI evaluation environments and benchmarks.

RF Optimization and KPI Analytics Engineer at Huawei (2025-11 – 2026-05)

• Monitor and analyze 4G/5G radio network performance KPIs to ensure optimal network quality and service reliability.
• Perform RF optimization activities including parameter tuning, troubleshooting coverage and interference issues, and improving network capacity.
• Analyze network performance data to identify anomalies and propose optimization strategies.
• Collaborate with cross-functional teams to enhance network efficiency and user experience.

End of Studies Internship - AI Research Intern at CESBIO, CNRS (2025-03 – 2025-08)

Contributed to the enhancement of the 3D DART model used for simulating satellite observations of land surfaces by integrating AI-based methods for atmospheric radiative transfer modeling. Supervised by Professor Jean-Philippe Gastellu

• Conducted a literature review and comparative analysis of analytical models simulating illumination and top-of-atmosphere signals across various spectral ranges.
• Designed and implemented an AI based approach to estimate variables for a new analytical model based on large-scale simulations of the DART radiative transfer model.
• Validated and integrated the improved analytical model into the operational DART system for public distribution.

Freelance - AI Developer at My Easy Transfer (2024-07 – 2024-09)

• Applied OCR technologies for card identification by leveraging Optical Character Recognition models to extract and verify information from identity cards.
• Implemented image verification using OpenCV and facial recognition to compare user video with ID images, improving authentication security.

Summer Internship - AI Intern at Knowlepsy (2024-07 – 2024-09)

• Performed feature engineering and applied clustering techniques to identify patterns and improve model performance for hand activity recognition.
• Utilized MLflow to track and manage machine learning experiments, ensuring efficient model development and reproducibility.

Summer Internship - Data Science Intern at My Easy Transfer (2023-07 – 2023-08)

• Performed data analysis to generate actionable insights supporting strategic decision-making and customer demand forecasting.
• Developed predictive models to forecast customer demand and improve model performance.
• Designed and deployed a chatbot using IBM Watson Assistant to enhance customer interaction and support services.

Degree in ICT Engineering (Information and Communication Technologies) – SUP'COM - Higher School of Communications of Tunis (2022-09 – 2025-10)

Program in Preparatory Cycle for Engineering Studies - Mathematics-Physics – IPEIT - Preparatory Institute for Engineering Studies of Tunis (2020-09 – 2022-07)

Diploma in Baccalauréat in Mathematics – Lycée Hannibal, Ariana (2016-09 – 2020-07)