Amadreza Farahani
Machine Learning Engineer @Zirak
Signup · Get unlimited contacts
WORK HISTORY
Machine Learning Engineer @Zirak
Led a team of two to ship a cross-platform LLM meeting assistant (Electron, React, FastAPI/Node.js) using RAG + LangChain/LangGraph with real-time ASR and speaker verification.• Built an internal LLM assistant (HuggingFace, vLLM/TorchServe) with MLflow/Grafana, reducing manual ops by ~50 hrs/month and enabling external deployment.• Developed a rail-inspection vision pipeline (PyTorch, YOLOv8) for defect detection and rail segmentation/tracking.• Boosted multispectral-image F1 by 10% via a custom UNet benchmarked against RF/XGBoost/KMeans; automated training/eval with Airflow.• Shipped a multispectral vegetation-detection pipeline (Docker, Kubernetes, SageMaker) with Prometheus metrics; cut manual review by 80% and compute cost by 35%(<3% drift).
EDUCATION
Iran University of Science and Technology
Bachelor's degree, Electrical, Electronics and Communications Engineering
Politecnico di Torino
Master of Science - MS, Computer Engineering
ABOUT AMADREZA FARAHANI
Machine Learning Engineer with a passion for developing AI solutions, I specialize in the end-to-end development of complex ML systems. Experienced in a wide range of technologies, including Python, PyTorch, TensorFlow, LLMs, and MLOps tools, and with a track record of boosting model accuracy, reducing resource usage, and deploying production-ready pipelines.I am always open to connecting with fellow professionals and exploring new opportunities to apply my skills to solve real-world challenges.
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.