Summary
Milad Barati is a Machine Learning Engineer with 8 years of experience building and deploying production-ready AI systems, currently focused on agentic and generative AI as well as ASR. At MCINext he led Persian Whisper fine-tuning to near state-of-the-art performance (24.51% WER), introduced PEFT for resource-constrained fine-tuning, migrated pipelines from Hugging Face to raw PyTorch to accelerate iteration 5×, and applied pruning/quantization to slash deployment costs. He bridges research and product, designing robust preprocessing/postprocessing and MLOps workflows while mentoring engineers to ensure real-world reliability. Comfortable across PyTorch, TensorFlow, ONNX and Kaldi, he also leverages LangChain/LangGraph for stateful RAG-powered agents. Currently pursuing an M.Sc. in AI at the University of Tehran, his research emphasizes trustworthiness and explainability—making agentic systems safer and more auditable in production.
8 years of coding experience
4 years of employment as a software developer
Master of Engineering - MEng, Artificial Intelligence, Master of Engineering - MEng, Artificial Intelligence at University of Tehran
Bachelor's degree, Computer Engineering, Bachelor's degree, Computer Engineering at University of Isfahan
High School Diploma, Mathematics, High School Diploma, Mathematics at National Organization for Development of Exceptional Talents (Sampad)