Mehdi Ziadi

Built a performance dashboard on AWS with CloudWatch and Grafana to improve incident response. Automated deployment using Terraform and AWS SDKs. Configured alerts via CloudWatch, SNS, SES, and Lambda to reduce downtime. Optimized monitoring costs by improving resource allocation.

• Built a multimodal AI tutor with LangChain for RAG, using LLMs for document queries and OCR text extraction. • Enhanced AI capabilities with web, video, and image search functionalities. • Designed responsive UIs in Next.js for better user interaction. • Fine-tuned Hugging Face models for accurate AI responses. Integrated APIs using FastAPI for efficient management of AI model interactions.

• Built and refined the "Clark" AI tutor using Ollama and the Mistral model with LangChain for RAG solutions, focusing on PDFs and document queries. • Designed and implemented responsive, intuitive user interfaces with Next.js. • Integrated and fine-tuned Hugging Face models and other services to ensure accurate and contextually relevant responses.