Ramsha S. Email & Phone Number

Fort Lauderdale, Fl, Us

Lahore, Punjab, Pakistan

-Fine-tuned Whisper model for realtime transcription, front end (client.js) sent chunks to backend (server.py) via web-sockets, backend then processed those chunks and returned transcription, deployed on AWS -Created a query-document using RAG for multiple docs which returns citation(s) of relevant docs on a query, used FAISS for vector embeddings, UI developed using streamlit which proved 90% accurate and reduced time to traverse the document by 85%-Created an LLM based Psychiatrist (prompt engineering), with a chat interface using streamlit, using Groq api key, streamed the output of LLM chain, integrated memory -Created a scraphgraph-ai for efficient web scraping which reduced time consumption by 75% -Created Docker images through dockerfile, used docker compose and deployed projects on AWS server-Created automated CI/CD pipeline to build, run and deploy

Lahore, Punjab, Pakistan

-Authentication & Security: Implemented JWT-based authentication for secure user login and registration, integrated CAPTCHA for bot protection, and enabled email verification and social login (e.g., Facebook OAuth).-Email & Communication: Configured MailChimp for automated newsletter management and developed an email backend for handling contact forms.-Payment Processing: Integrated Stripe API for handling subscription-based premium access with secure payment transactions.-Access Control: Developed CRUD operations with Django, applying role-based access control (RBAC) with permissions for editor, creator, and admin roles.-Engagement Features: Designed and implemented a scalable commenting system with like functionality, real-time notifications using WebSockets, and threaded replies.-Real-Time Communication: Established WebSocket connections for real-time user-specific notifications and broadcasted updates to all connected clients (validated with WebSocket King).-Machine Learning Integration: Integrated ML models into the backend to leverage predictive analytics and enhance application intelligence.-Performance Optimization: Utilized database caching strategies, implemented GraphQL for optimized querying, and enabled data export in Excel format. Configured browser push notifications and integrated a robust search functionality.-Media Handling: Developed an image upload system with efficient processing and management.-Docker Containerization: Employed Docker for containerizing the application, ensuring environment consistency, simplifying deployment, and facilitating scalable infrastructure.-Frontend Integration: Integrated backend APIs with ReactJS and VueJS frontends for a seamless and dynamic user experience.

-Web scraping of 10,000+ records through selenium automation which resulted in 60% time reduction-Data formatting through Microsoft Excel-Image cleaning through automated python script, which proved 98% accurate and reduced time taken by 50%

Lahore, Punjab, Pakistan

-Training of YOLOv8 for player detection, data-set created using ffmpeg through shell, data annotations in label-studio-Data-set preperation of football-detection model, automating annotations through label studio which resulted in time reduction by 70%-Data-set creation for sticker generation model, reduced time consumption by 50%-Web scraping and downloading of 30,000+ records using selenium automation

Bachelor'S Degree, Computer Science

A Levels, Engineering

Activities and Societies: Photography Club Badminton team