Zakaria Laabsi@zakaria_laabsi249

I contributed to a cybersecurity convention, presenting a poisoned CNN model to demonstrate AI vulnerability with adversarial backdoor attack, thereby underlining the critical need for robust models.

This project aims to create an engaging AI English Tutor, combining the state-of-the-art natural language processing capabilities of OpenAI's GPT-3.5-Turbo model with ElevenLabs's high-quality text-to-speech technology, all presented in an intuitive, accessible Streamlit interface. The tutor offers efficient learning methods to enhance English fluency, correcting users' English sentences and initiating dialogues for practice. Through the OpenAI's model, the tutor generates real-time responses to user queries and provides corrections to improve English skills. It then uses ElevenLabs's technology to generate audio responses, providing auditory reinforcement to the learning experience. The project is implemented as a Streamlit application, providing a web-based front-end that allows users to easily interact with the AI tutor. The application requests English sentences from the user, processes them with GPT-3.5-Turbo, and vocalizes the responses using ElevenLabs's API. Users have the ability to select different voices for the output, enhancing the personalized learning experience. In terms of deployment, the application uses GitHub Actions for CI/CD, allowing for continuous updates and seamless deployment. API keys are securely stored as GitHub Secrets, maintaining the security of sensitive data. Overall, this project serves as a showcase of how AI technologies can be integrated to create a comprehensive learning tool, and how they can be made accessible through intuitive user interfaces.

Microtargeting Solution for Performance & Growth Marketers and Ad Specialists Develop a comprehensive microtargeting solution that leverages data-driven insights to empower performance and growth marketers and ad specialists to create precise and high-converting ad campaigns. Key Features: - User-Defined Product/Topic Input - Dynamic Research & Listening Engine - Database Integration & Insight Extraction - Product detailed, brand guidelines and custom branded graphic input - Ad & Landing Page Generation Tool For the solution we utilized the AI Agents, based on SuperAGI, Langchain, and GPT4 API.

GenieFront represents the next frontier in web development, merging the worlds of design and coding. Have you ever created an exquisite design on platforms like Figma and wished for an immediate transition to a functioning website? Enter GenieFront. With the integration of Azure Computer Vision, Keras-OCR, and the unparalleled capabilities of GPT-4, our tool translates intricate designs into dynamic web content, complete with finely-tuned HTML and CSS. Targeting everyone from individual freelancers to large enterprises, GenieFront aims to revolutionize the web development process, making it swifter, more intuitive, and impeccably accurate. Join us in redefining the digital landscape.