Tinsae Abreham

Tinsae Abreham SUMMARY Passionate Machine Learning and Flutter developer with a BSc in Electromechanical Engineering, seeking internship opportunities to apply expertise in computer vision, NLP, and automation. Proficient in TensorFlow, PyTorch, and OpenCV, with hands-on experience in real-time object detection and sentiment analysis projects deployed on Hugging Face. EDUCATION Addis Ababa Science and Technology university Addis Ababa, Ethiopia BSc in Electromechanical Engineering June 2021 - June 2025 SKILLS SUMMARY ● Languages: Python, Flutter, C++ ● Frameworks: NumPy, scikit-learn, TensorFlow, Keras, PyTorch, Flask, Hugging Face ● Tools: Power BI, Excel, Tableau, Matplotlib, Pandas, Seaborn, Git, OpenCV, Plotly ● Platforms: PyCharm, Jupyter Notebook, Visual Studio Code, Google Colab, Kaggel ● AI/ML: Computer Vision, NLP, Generative AI, Model Evaluation, Model Deployment, Transfer Learning WORK EXPERIENCE AUTOMATION ENGINEER INTERN | ROBOX Jun 2024 - Sep 2024 ▪ Designed mechanical components using SolidWorks and led training sessions to upskill peers in CAD design and product modeling. ▪ Mentored students in robotics principles through hands-on, team-based projects. ▪ Co-developed and simulated an injera-baking robotic arm, contributing to mechanical design, sensor integration, and motion planning for real-world application in local food automation. AUTOMATION ENGINEER INTERN | iCog_Labs Jun 2025 - Present ▪ Contributed to AI research in the Pattern Miner team, focusing on pattern mining and knowledge representation. ▪ Developed and tested algorithms using MeTTa for structured data processing. ▪ Researched and tested new AI model PROJECTS Bird Detection & Tracking | LINK May 5 – May 11 2025 ▪ Achieved 80%+ tracking accuracy with entry time and movement logging for each bird ID ▪ Built a real-time system to detect and track birds in image/video using YOLOv11 and Deep SORT ▪ Deployed a live demo with interactive UI using Gradio on Hugging Face Spaces X Sentiment Analysis | LINK Mar 23- Apr 3 2025 ▪ Fine-tuned a pre-trained BERT model on a dataset of 900,000 text samples for binary sentiment classification. ▪ Achieved 88% accuracy on the validation set, outperforming traditional machine learning baselines. ▪ Conducted error analysis to improve model performance and interpret misclassified instances. ▪ Deployed model on Hugging Face for real-world sentiment analysis. Transformer Model Implementation | LINK Mar 2 – Mar 5 2025 ▪ Designed the project for educational use to demonstrate step-by-step workings of the Transformer architecture. ▪ Built a Transformer model from scratch using Tensorflow based on the “Attention Is All You Need” paper. ▪ Demonstrated the model with sample data to visualize step-by-step processing and attention mechanisms. CERTIFICATES Advanced Computer vision With Tensorflow | CERTIFICATE Mar 2025 ▪ Mastered deep learning techniques for computer vision, including image classification, segmentation, and object detection. ▪ Explored image classification, segmentation, object localization, and detection. ▪ Applied transfer learning and built custom object detection models (e.g., R-CNN, ResNet-50). ▪ Used class activation and saliency maps for model interpretation and analysis. Tensorflow Developer | CERTIFICATE Feb 2025 ▪ Mastered best practices for TensorFlow to train neural networks for computer vision and natural language processing. ▪ Built NLP systems using TensorFlow, implementing RNNs, GRUs, and LSTMs. ▪ Applied advanced techniques to train models using large text repositories for natural language tasks. Machine Learning with Python | CERTIFICATE May 2024 ▪ Mastered building ML models with NumPy and scikit-learn for supervised tasks, including linear/logistic regression and decision trees. ▪ Trained neural networks with TensorFlow for multi-class classification and applied tree ensemble methods. ▪ Built recommender systems using collaborative filtering and deep learning, as well as a deep reinforcement learning model.