Raaid Yousuf

AI Engineer specializing in Generative AI, Agentic Workflows, and LLM Applications.

Based in: Islamabad, PakistanMost recently: AI Intern, Xavinex

Software Engineering graduate with expertise in building multi-agent systems, RAG pipelines, and high-accuracy computer vision models. Proven track record in developing end-to-end intelligent automation solutions and scalable AI-powered platforms.

Experience

AI Intern

XavinexLahore - PakistaninternshipJul 2026 – Aug 2026

Summary

Benchmarked and optimized SOTA vision models (Swin, MobileViT, YOLO & EfficientNet) for garment defect detection, achieving 95% accuracy. Optimized model performance through dataset refinement, taxonomy optimization, and error analysis. Automated RCA & CAPA generation using LLMs for quality management.

Results

Achieved 95% accuracy in garment defect detection using SOTA vision models.

SwinMobileViTYOLOEfficientNetLLMsComputer Vision

Data Science Intern

Kangaroo VenturesUnited States (Remote)internshipAug 2025 – Sep 2025

Summary

Conducted data preprocessing, exploratory data analysis (EDA), and feature engineering for ML pipelines, improving dataset quality and model performance. Assisted in model training, evaluation and hyperparameter tuning using Python ML frameworks. Supported research and experimentation workflows, reducing validation time by 5%.

What I did

Optimized model performance by adding more useful features.

Results

Reduced validation time by 5% through supported research and experimentation workflows.
Processed approximately 1 million textual data records and 1 year of time series data.

PythonML frameworksEDAFeature Engineering

Skills

Technical

RAG

NLP

Computer Vision

Python

LangChain

Prompt Engineering

LlamaIndex

PyTorch

TensorFlow

Scikit-learn

Embeddings

Fine-Tuning

Hugging Face

Transformers

OpenAI / Groq API

LangGraph

Vector Databases

Pandas

Git

GitHub

Google Colab

VS Code

Matplotlib

Seaborn

Plotly

C++

Java

FAISS

SQL

JavaScript

Power BI

React

Next.js

Streamlit

Projects

AI-Powered Cheating Detection System

Lead Developerpersonal

Built a real-time computer vision system using Python, OpenCV, and YOLO-based object detection to monitor exam environments and detect suspicious behaviors. Integrated gaze tracking, face detection, and object detection, achieving 89% detection accuracy on a labeled dataset of 10k images.

PythonOpenCVYOLOComputer Vision

Multi-Agent Research System

Lead Developerpersonal

Built an automated multi-agent pipeline using LangChain, Gemini 2.5 Flash, and Python, integrating the Tavily API and BeautifulSoup for precise web searching and data extraction. Developed a responsive Streamlit platform to orchestrate a 4-stage workflow (Search, Read, Write, Critic), efficiently parsing up to 3,000 characters per source to generate and autonomously evaluate structured reports.

LangChainGemini 2.5 FlashPythonTavily APIBeautifulSoupStreamlitGitHub

AI-SmartNaggar

Lead Developerpersonal

Built AI SmartNaggar, a multi-modal AI-powered civic issue reporting system enabling citizens to report urban problems via text, image, and voice, using MobileNetV2 (PyTorch) for image classification, Whisper for speech-to-text, and TF-IDF + Logistic Regression for NLP-based complaint categorization. Developed a Streamlit + Supabase platform with Folium geolocation mapping that automatically routes complaints to relevant municipal departments, enabling real-time tracking, analytics, and faster resolution of civic issues.

MobileNetV2PyTorchWhisperTF-IDFLogistic RegressionNLPStreamlitSupabaseFoliumGitHub

AI Imaginator

Frontend Developerpersonal

A full platform for image generation and editing including text-to-image, face swap, and in-painting, featuring a real-time community messaging and content sharing system.

ReactNext.jsNode.jsExpress.js

Toxic Comment Classifier

personal

Built a machine learning model to classify toxic comments.

Heart Disease Prediction

personal

Trained a machine learning model for predicting heart disease.

Sentiment Analysis

personal

Developed and trained a machine learning model for sentiment analysis.

Education

Bachelor of Science

University of Engineering & Technology, TaxilaSoftware EngineeringGPA 3.8Sep 2022 – Jul 2026

Recognition

Awards

4th Position in Hackathon

Universities in Pakistan

Ranked 4th across all universities in Pakistan for the AI-SmartNaggar project during a hackathon.

Certifications

Google AI Professional Certificate

Google2026

NLP-RAG Systems & AI Agents

Awfera2026

IBM Data Science Professional Certificate

IBM2025

Career Essentials in Software Development

Microsoft & LinkedIn2024

AI agents using typescript

Not specified

Hire RaaidReach out about a role, a contract or a conversation.For recruiters

Raaid's twin is AI, it can make mistakes.

More people on Kanvis

Browse more people Create your own Kanvis