National Engineering Diploma in Applied Mathematics and Modeling

Generative AI & ML Engineer

Developed AI agents and multi-agent systems in Python using LangChain and LangGraph
frameworks, incorporating prompt engineering with LLMs (OpenAI, Gemini, Groq, etc.).
• Automated complex processes by creating example agents—such as a customer support agent
that leverages RAG techniques with vector stores and embeddings for enhanced text
generation and context retrieval, and a Calendly agent for managing daily tasks using Google
Calendar APIs.
• Managed data with PostgreSQL and integrated multi-agent systems using Neo4j for scalable,
graph-based database solutions.
• Contributed to building a no-code platform for creating and deploying AI agents and
multi-agent workflows, enabling users to automate complex processes and drive business
impact without writing code.
• Built and deployed Django-based RESTful APIs to facilitate seamless communication between
AI components and external services.
• Developed a real-time Voice Assistant:
• Integrated the OpenAI API with Voice Activity Detection (VAD) for real-time speech
conversion.
• Implemented an open-source local solution on a server, combining Kokoro TTS for speech
synthesis and HuggingFace Whisper STT for transcription.
• Fine-tuned a text-to-speech (TTS) model for the Tunisian dialect by collecting, preprocessing,
and training on Tunisian voice data to enhance speech synthesis quality and naturalness.
• Deployed a TTS model for real-time voice bot interaction on AWS using S3 and SageMaker to
compare the fine-tuned open-source model for Tunisian dialect with an OpenAI API-based
model