Hossein Khalilian
Tehran, Iran ยท
hse.khalilian08@gmail.com
About Me
I am an AI Engineer with a Master's degree in Artificial Intelligence from Amirkabir University of Technology and over 5 years of specialized experience in machine learning, natural language processing (NLP), and scalable AI solutions. I am proficient in developing and optimizing large language models (LLMs) such as GPT and BERT, as well as implementing advanced NLP techniques, including Named Entity Recognition, Machine Translation, and Text Summarization. My technical expertise spans Python, PyTorch, FastAPI, Docker, and various databases like vector and graph databases.
At Mirror Group, I enhanced Persian speech recognition models, reduced WER by 28%, and led NLP projects that boosted content retrieval accuracy and processing efficiency by over 40%. My work in Mofid Securities focused on anomaly detection, knowledge distillation, and deploying models in production environments, achieving up to 88% latency reduction. I have a proven track record of scaling machine learning projects through containerization and API integration, contributing to increased operational efficiencies.
I'm passionate about leveraging AI to solve real-world challenges and consistently deliver impactful results. Always eager to learn, I stay updated with the latest developments in AI and regularly participate in conferences and collaborative initiatives to advance the field. Feel free to connect if you're interested in my work or wish to discuss potential collaborations.
Experience
AI Engineer
Mom Fertility Center
Responsibilities:
- Significantly enhanced speech recognition models for Persian, achieving a 28% performance improvement and reducing WER to 7.4%.
- Developed and optimized NLP models for text classification, summarization, and translation using Hugging Face Transformers, improving text retrieval accuracy and processing efficiency.
- Engineered recommendation systems increasing website traffic by 70%, and implemented efficient product matching for 6 million products, reducing processing time to 28ms.
- Deployed scalable machine learning solutions, including API development and containerization with Docker, enhancing deployment efficiency by 40% and labeling efficiency by 25%.
- Utilized prompt engineering with LLMs for high-quality text generation, streamlining content creation processes.
Technologies:
-
Python, PyTorch, Huggingface Transformers, spaCy, Vector
Databases, FastAPI, gRPC, Docker, Label Studio, Git, Prompt
Engineering, Large Language Models, LangChain, Full-text Search
Nov 2021 - Present
Machine Learning Engineer
Mofid Securities
Responsibilities:
- Gathered, cleaned, and preprocessed data for training machine learning models, ensuring high-quality datasets for model development.
- Designed and trained machine learning models for anomaly detection in images using knowledge distillation techniques, achieving an accuracy of 92% and an improvement of 26% over previous models.
- Optimized model operations by converting and compressing trained models for improved inference time on CPUs, achieving an 88% reduction in latency.
- Dockerized projects and designed APIs for model integration, enhancing system scalability and accessibility.
Technologies:
-
Python, Docker, API Development, Knowledge Distillation
Techniques, Model Compression
Apr 2021 - Sep 2021
Education
M.Sc. Computer Engineering - Artificial Intelligence
Amirkabir University of Technology
Specialized in Artificial Intelligence with a thesis on
Distributed Deep Learning. Achieved 4th rank among 53 students in
the program.
2018-2021
B.Sc. Information Technology (Double Major)
Isfahan University of Technology
Graduated as the top student, 1st rank among 83 students.
Completed a significant project on the Design and Implementation
of a VoIP Gateway Device.
2014-2018
B.Sc. Electrical Engineering
Isfahan University of Technology
Graduated in the top 10% of the class among 120 students. Gained
comprehensive knowledge in electrical engineering principles and
their applications.
2013-2018
Skills
Programming Languages & Tools
Data Science Toolkit
- Python, Golang, Bash
- Machine Learning, Deep Learning: Numpy, Pandas, Scikit-Learn, PyTorch, Matplotlib, Plotly, FastAPI, Langchain
- NLP: Large Language Models, Prompt Engineering, Retrieval-Augmented Generation(RAG), Hugging Face Transformers, NLTK, Spark NLP, Spacy
- AI: Automatic Speech Recognition, Recommender Systems, Reinforcement Learning
- Version Control Systems: Git
- Soft Skills: Problem Solving, Team Working
Personal Projects
Solving Placement Problem for Deep Neural Networks
Detection of Earth Dam Breaching by image processing techniques
