Mohamed Rayen Khlifi

-Fine-tuned an LLM assistant on Oodrive datasets to automate OpenAPI validation, improving accuracy by 25%
and reducing manual linting by 40%.
-Built an LLM system to generate Oodrive-compliant YAML and Javascript rules, cutting rule creation time by
60%.
-Designed a Mistral-driven tool for Postman test scenario generation, automatically covering 80% of test cases.
-Deployed AI assistants via FastAPI into CI/CD pipelines, reducing deployment time and improving automation

-Developed a computer vision system for leaf recognition and classification for biomedical purposes, based on a dataset of 15, 000 annotated images.

-Implemented and optimized CNN models with TensorFlow and Keras, achieving 91% classification accuracy while improving inference speed.

-Developed an end-to-end application using Flask for the backend and React for the frontend.

Collaborated on global tech challenges and delivered AI ethics presentations, strengthening leadership and public
speaking skills.

-Developed and executed black-box tests and automated CI/CD pipelines with Ranorex and Jenkins, reducing
testing and deployment time by 30%, and managed test binaries using Artifactory and Bitbucket to improve
version control and traceability.
-Identified and resolved critical bugs, enhancing software reliability and reducing post-release incidents by 20%