cv
General Information
| Full Name | Farid Kazimov |
| Title | Master Engineer | AI Engineer | Computer Vision | LLM & GenAi | ML |
| kazimov.ferid.99@gmail.com | |
| Languages | Azerbaijan (Native), Turkish (Native), English (Fluent) |
Education
-
2023 - 2025 Master of Science (M.Sc.)
Wroclaw University of Science and Technology, Wroclaw, Poland - Applied Computer Science.
- Specializing in Computer Vision and Deep Learning.
-
2017 - 2021 Bachelor of Science (B.Sc.)
Kahramanmaras Sutcu Imam University, Kahramanmaras, Turkey - Computer Engineering.
Experience
-
June 2021 - July 2021 Software Engineer Intern
Mavi Yesil Yazilim ve Otomasyon - Assisted in industrial automation solutions and system integration.
- Performed software validation and contributed to process control systems.
-
June 2019 - July 2019 Network Engineer Intern
KSU IT Department - Participated in network setup, configuration, and troubleshooting.
- Prepared technical documentation for IT infrastructure maintenance.
Projects
-
2024 Automatic Animal and Crowd Counting
- Developed a computer vision system using YOLOv8 with heatmap and density maps for counting animals and people in dense environments. Validated on datasets including sheep, bees, cows, monkeys, butterflies, and humans.
- YOLOv8 • OpenCV • Heatmap and Density Maps • Computer Vision
-
2024 Transfer Learning for Crowded Animals Detection
- Applied transfer learning to adapt YOLOv8 models trained on one species to others (e.g., cows, monkeys, butterflies). Cross-class fine-tuning improved generalization and reduced training data requirements for scalable deployment.
- Transfer Learning • Fine-Tuning • Cross-Class Generalization • Domain Adaptation
-
2025 YOLO Version Comparison (YOLOv5 -> YOLO12)
- Benchmarking YOLO models for detecting animals in crowded herds, focusing on accuracy, inference speed, occlusion robustness, and cross-species generalization.
- YOLOv5–YOLOv12 • Backbone Comparison • Anchor-Free Detection • Non-Maximum Suppression (NMS) • Inference Speed and Latency • Model Optimization (Quantization, Pruning)
-
2025 Smart Research Assistant (Langgraph AI Agent)
- Autonomous AI agent using LangGraph/LangChain with tool-use + web search; deployed on Streamlit.
-
2025 Document QA System (RAG Project)
- Single-document QA system using local embeddings (MiniLM) and FAISS vector search.
-
2025 AI-Powered Traffic Analysis System
- Real-time traffic analysis using RT-DETR detection, tracking, speed estimation, and red-light violation logging, with CSV/JSON outputs and an LLM-generated summary.
Skills
-
LLM & GENAI
- Prompt Engineering • Function Calling • RAG • LangChain • LangGraph • AI Agents • Embeddings • Hallucination Reduction • Safety Filters • Azure/OpenAI • HF Transformers • NLP (classification, summarization, QnA)
-
AZURE AI (AI-102 CAPABILITIES)
- Cognitive Search • AI Language • Document Intelligence (OCR) • Speech & Vision Services • Content Safety • Prompt Flow • Azure ML (deployment)
-
MACHINE LEARNING AND DEEP LEARNING
- Neural Networks • CNNs • Transformers • Data preprocessing • Regularization(Dropout, BatchNorm) • Evaluation (F1, mAP, RMSE) • Explainability (SHAP, LIME) • Hyperparametertuning
-
COMPUTER VISION
- YOLO • OpenCV • Heatmaps/Density Maps • Object Detection • Segmentation • Transfer Learning • Cross-species generalization • Tracking & Occlusion Handling
-
INFORMATION RETRIEVAL AND VECTOR SEARCH
- FAISS • Semantic/Hybrid Search • Similarity Search • QueryExpansion • Embedding Optimization
-
MLOPS & DEPLOYMENT
- Docker • GitHubActions • CI/CD • Model versioning • Batch/real-time inference • Azure deployment • Monitoring & Logging
-
PROGRAMMING & TOOLS
- Python • C++ • FastAPI • NumPy • Pandas • Scikit-Learn • PyTorch • TensorFlow • Keras • Git • Colab • VSCode
Other Interests
- Reading technical papers, Hiking, Learning new programming frameworks.