Abdullateef Muadh Olamilekan

Data Scientist & Analyst /Computer Vision/Machine Learning Engineer https://www.linkedin.com/in/muadh-abdullateef/ https://github.com/freshiwe ABDULLATEEF MUADH OLAMILEKAN Short Bio A young, dynamic, result-oriented, and productive Individual. I am a fast learner and an open-minded team player who is passionate about Data, with the hope of making significant contributions through innovative research. Interests: Data Analysis and Data Science, Machine Learning, Deep Learning, Computer Vision and Large Language Models (LLMs) Education BACHELOR OF TECHNOLOGY – KIIT University – India Majors: Computer Science and Communications Engineering BACHELOR OF ART – International Open University Majors: Arabic Language and Linguistics September 2021 - 2025 September 2019 - 2024 Skills • • • • C/C++ Python (Pandas, NumPy, SciPy, matplotlib, Tensorflow, Seaborn) OpenCV Yolo • • • • Model Deployment(Django,Flask) Power BI, Tableau Pytorch Smolagent, llanggraph, n8n, CrewAI, llangchain,e.t.c. Core Expertise  Deep Learning & Neural Networks (TensorFlow, Keras, PyTorch)  Natural Language Processing & Recommender systems            Computer Vision & Image Processing (object detection, image segmentation, openCV) Data cleaning, wrangling, and analysis (pandas, numpy, sci-kit-learn) Data Mining & Web Scraping (selenium, beautiful soup, scrappy) Data Science Apps (Flask) Dashboards and Reports Technical Writing (blogs, articles, tutorials ) Project Reports (proposals, plans, QA, user manuals) Business Intelligence (Tableau, PowerBI) Visual & Descriptive analysis (Matplotlib, seaborn, plotly) Finetuning Generative AI models. Building AI Agents for automating Industrial workflows. Few Projects executed FACE DETECTION SYSTEM ● I developed a robust face detection system capable of real-time processing using OpenCV and deep learning algorithms. ● I achieved high detection accuracy by implementing state-of-the-art models such as Haar cascades and deep neural networks, ensuring reliable performance across various lighting conditions and angles. FACE RECOGNITION MODEL ● ● I implemented a face recognition system that accurately identifies individuals by leveraging convolutional neural networks (CNNs) and deep learning techniques for both image and video data. I ensured the system's scalability to handle large datasets and multiple users, optimizing the model for both speed and accuracy. OBJECT TRACKING SYSTEM ● ● I developed an object-tracking system using KCF (Kernelized Correlation Filters) and CSRT (Discriminative Correlation Filter with Channel and Spatial Reliability) algorithms to enhance tracking accuracy and robustness in various scenarios. I optimized the system to achieve real-time tracking performance, handling multiple objects simultaneously with high precision in diverse and dynamic environments. REAL- TIME FACE MASK DETECTOR ● ● Implemented an efficient face mask recognition algorithm, demonstrating proficiency in computer vision and pattern recognition. Implemented an efficient face mask recognition algorithm, demonstrating proficiency in computer vision and pattern recognition. SOCIAL DISTANCING DETECTOR ● ● Implemented a computer vision system for detecting and monitoring social distancing violations, showcasing expertise in leveraging advanced technologies for public health solutions. Developed real-time monitoring capabilities, allowing instant identification of social distancing breaches and enabling timely interventions. OBJECT DETECTION/ OPTICAL CHARACTER RECOGNITION ● ● Implemented cutting-edge techniques for object detection and optical character recognition, showcasing a forwardthinking approach. Achieved high accuracy rates for both object detection and optical character recognition, demonstrating the effectiveness of the implemented solutions. OBJECT DETECTION WITH YOLO ● ● I implemented an object detection system using the YOLO (You Only Look Once) algorithm, achieving high detection speed and accuracy for real-time applications. I fine-tuned the YOLO model to accurately detect and classify multiple object types in various environments, ensuring robustness and reliability in diverse conditions. STYLE TRANSFER USING DEEP LEARNING ● ● I implemented a neural style transfer system to blend artistic styles with content images, creating visually appealing artwork by leveraging deep learning techniques. I optimized the style transfer model for improved performance and faster processing times, enabling high-quality results in a variety of artistic styles. IMAGE SEGMENTATION USING MASK R- CNN ● ● I developed an image segmentation system to accurately segment objects and regions from both static images and video frames, utilizing advanced techniques such as U-Net or Mask R-CNN. I adapted the segmentation model to handle various types of input data, ensuring high-quality results across different environments and applications, including dynamic video scenes. ASL RECOGNITION WITH DEEP LEARNING ● Developed a convolutional neural network (CNN) using deep learning techniques to classify images of letters from  American Sign Language (ASL). Demonstrated expertise in computer vision and neural network architectures, achieving accurate and reliable classification results for ASL gestures. GIVE LIFE: PREDICT BLOOD DONATIONS   Developed a binary classifier using machine learning techniques to predict the likelihood of a blood donor donating again. Applied feature engineering and model optimization to enhance prediction accuracy, showcasing a data-driven approach to address a critical healthcare challenge. PREDICTIVE MODELING FOR AGRICULTURE ● ● Delved into the field of agriculture, applying supervised machine learning techniques and feature selection to address practical challenges in crop cultivation. Demonstrated proficiency in leveraging machine learning for precision agriculture, providing valuable insights to optimize crop yield and resource allocation. COMPARING COSMETICS BY INGREDIENTS ● ● Processed and analyzed ingredient lists for cosmetics on Sephora, demonstrating expertise in data preprocessing and extraction. Utilized t-distributed Stochastic Neighbor Embedding (t-SNE) and Bokeh for visualization to represent the similarity between cosmetic products based on their ingredients. LLM- BASED PROJECTS ● Context-aware book recommendation system. ● Automated Email sorting agent using Langgraph. ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● Implementation of Customer Outreach Campaign AI Agent using CrewAI. Implementation of Vision Agent that using Langgraph. Implementation of Agentic RAG based project using LLamaindex for answering questions based on needs. Implementation of AI agent for blog writing using CrewAI. Two way AI Email agent using n8n. Multiagent framework implementation using Smolagent. Fine-tune LLM with the LORA model for text classification. Fine-tuning the DistilBert model for intent recognition. Building a chatbot model for text generation. Fine-tuning LLM model for machine language translation. Multiclass classification of news categories by fine-tuning of TFBert Model. Paraphrase generation by fine-tuning the T5 model. RAG implementation for elastic search. Implementation of long document classification for text classification using Longformer. Text summarization by fine-tuning the LLM model Fine-tuning LLM Model for YouTube video recommendations. Work/Research Experience Digital Marketing Assistant- Upwork Oasis Infobyte – Data Science Intern The Spark Foundation – Data Science Intern Certifications Google UI/ UX Design - Coursera Google IT SUPPORT – Coursera Google Data Analytics – Coursera Virtual Experience Program – TATA Data Science with Python – Data camp March 2021 - October 2022 October – November 2022 September – October 2023 Computer Vision - Udemy Machine Learning Track – Data camp Applied Data Science Lab – World Quant University Applied AI Lab – World Quant University Data Engineering – Data camp Research Papers.  • • A Comparative Study on Sentiment-Aware Multimodal Personalized Movie Recommender System. In Proceedings of the International Conference on Data Analytics and Insights (ICDAI-2024). https://doi.org/10.1007/-_36 Web Service Selection Using Hierarchical Graph Transformation. A CNN-BERT multimodal approach towards improving Web Service Classification.