Data Scientist | Machine Learning Engineer
📍 Currently in Berlin, Germany

I am a passionate Machine Learning Engineer and Data Scientist with over 2 years of experience. My expertise spans predictive modeling, Computer Vision, Generative AI, and Natural Language Processing. I specialize in solving complex challenges using state-of-the-art technologies like Deep Learning, Transformers, and Contrastive Learning.

🔍 I thrive in managing end-to-end machine learning projects, including data collection, model deployment, and performance monitoring.

• Python (5 years) | R | C++

• Computer Vision: YoloV5-V8, Faster R-CNN, SAM (Segment Anything Models), Siamese Networks
• Generative AI: GANs, WGAN, CycleGAN, Stable Diffusion, Kadinsky
• NLP: BERT, HF Transformers, spaCy, NLTK
• Frameworks: PyTorch, TensorFlow, Scikit-Learn

• Git, Docker, Azure ML, Linux

• Pandas, NumPy, PowerBI, Matplotlib

• Evaluated and implemented Generative AI models (Stable Diffusion, ControlNet) for synthetic data generation in industrial applications.
• Developed solutions using state-of-the-art Computer Vision tools for image segmentation and similarity detection.
• Tools & Technologies: PyTorch, TensorFlow, FastSAM, MobileSAM.

• Developed a system for analyzing vibratory patterns in honey bees using Computer Vision and statistical methods.
• Tools & Technologies: Python, Deep Learning, OpenCV.

• Built a system to predict optimal travel routes using historical and real-time traffic data.
• Technologies Used: Google Directions API, TensorFlow, Random Forests.

• M.Sc. in Data Science
  Freie Universität Berlin (FU), Germany

  • Focus: Generative AI, Computer Vision, Deep Learning, Digital Economy.

• B.Sc. in Computer & Information Science
  PMAS-UAAR, Pakistan

  • Focus: Software Engineering, Artificial Intelligence.

• Email: mkaleemcs@gmail.com
• GitHub: github.com/Muhammad-Kaleem-Ullah
• LinkedIn: linkedin.com/in/muhammad-kaleem-ullah

• Würth Group, Berlin, Germany (2023 - Present)

  • Developed advanced Computer Vision solutions using models like YoloV8, FastSAM, and Contrastive Learning.
  • Created and deployed tools for image segmentation, classification, and localization.

• Unzer GmbH, Berlin, Germany (2022)

  • Built transaction risk classification models using Gradient Boosting Machines and Neural Networks.

• PIEAS, Islamabad, Pakistan (2020)
  • Developed trip-planning systems based on traffic prediction using Google APIs.

When I'm not working with AI models, I enjoy exploring new technologies, contributing to open-source projects, and engaging in research at the intersection of Generative AI and Computer Vision.

💡 Let's connect and build innovative solutions together!