This app uses AI to explain complex concepts in Urdu script with English scientific terms, leveraging relatable examples from daily life in Pakistan. Students can even ask questions via text or microphone, and get instant, context-aware answers. It's rewarding to build tech that makes learning more accessible and engaging!
App Link: AI Science Teacher
Implemented a paid Live Chat box for websites — powered by AI and Python — that provides automated responses from a custom document-fed dataset.
Built and deployed using Replit, this smart chat solution can instantly answer user queries based on uploaded documents — making website support faster and more efficient.
By providing the sender's email and its content, it shows sensitivity and suggests the email and its content. It also tells whether either open these links or not.
App Link: Phishing Attack Detector
YouTube Link: Presentation
Github: Link
As part of the project, I also integrated an email verifier module to ensure data integrity. This verifier checks the validity of email addresses through multiple layers: format validation, domain verification (MX records), and SMTP-level mailbox checks — all without sending an actual email. This helps filter out invalid, disposable, or role-based addresses, improving overall reliability and trustworthiness of the system.
I’m excited to share the successful completion of my recent project focused on AI-generated content detection, powered by Agentic Retrieval-Augmented Generation (RAG) and Python.
A smart detection system capable of identifying AI-generated text using the power of agentic workflows and custom retrieval pipelines.
Imbeded in sopercart.com
It works with an embedded Data Set file and checks by image availability. Link
It works with an embedded PDF file and answers through it. link
It stores the voice parameters and generates voice according to the new text's predefined parameters. link (Private).