Publication Profile
Google Scholar
Educational Details
- Master of Applied Science in Electrical and Computer Engineering (Sep 2020 – Oct 2022)
University of British Columbia, Vancouver, BC, Canada
- Bachelor of Science, Dual Major in Electrical Engineering and Computer Science (Sep 2014 – Jul 2019)
Sharif University of Technology, Tehran, Iran
Professional Experience
Mohammad Amin Roohi is a skilled software and machine learning engineer with over five years of experience in developing advanced AI systems. He is the co-founder of Contacti AI, where he leads the design and implementation of conversational AI systems powered by large language models (LLMs), ultra-natural text-to-speech (TTS), and speech-to-text (STT) technologies. He has expertise in Retrieval-Augmented Generation (RAG), NLP pipelines, and AI-driven task automation using Mixture-of-Experts (MoE) models.
In his previous roles, he optimized computer vision models for edge devices at Visionish AI and contributed to research at UBC, creating a mathematical model for predictive quality control in composite manufacturing. Mohammad Amin has also built robust data pipelines, full-stack solutions using Django, and end-to-end CI/CD processes with Docker and GitHub Actions.
Research Interest
Mohammad Amin’s research focuses on machine learning, deep learning, and AI-driven automation. His expertise spans NLP, computer vision, and advanced optimization techniques, including attention mechanisms, domain adaptation, and transformers. He is particularly interested in leveraging AI to solve practical problems, such as conversational AI, composite manufacturing quality control, and real-time wildlife detection systems.
Author Metrics
- Publications:
- Safe Optimization with Black-Box Information: Application to Composites Autoclave Processing
- Safe Away Step Frank-Wolfe Algorithm under Uncertain Constraints
- Awards:
- Secured 2nd place in the Beeloud and Build AI 2024 Hackathon ($1k prize)
- $12k NFRF project grant for applied research methods (Feb 2021)
- Achievements:
- Developed an open-source Python library for safe optimization algorithms.
- Recognized for innovative AI applications, reducing costs and improving quality in industrial environments.
Publication Top Notes
- Title: Safe Optimization Developments and Applications
- Author: Mohammad Amin Roohi
- Institution: University of British Columbia
- Year of Publication: 2022
Conclusion
Mr. Mohammad Amin Roohi demonstrates exceptional suitability for the Best Researcher Award based on his innovative contributions to optimization, machine learning, and AI. His ability to address real-world challenges through cutting-edge research and practical applications makes him a strong contender for this recognition. Expanding his publication record, industry collaborations, and community engagement can further solidify his standing as a leading researcher. Overall, his work reflects both depth and breadth, aligning well with the objectives of the award.