Amin Katebsaber

• Developed a real-time policy processing system that accelerated policy issuance by 20% and improved accuracy by 90%. • Built a self-service system for automated automobile damage assessment and repair cost estimation, enabling a 30% reduction in claim processing time. • Created an LLM-powered agentic system that handles user inquiries, leading to a 50% reduction in call center load during peak demand.

• Established the data science division, providing comprehensive AI as a Service (AIaaS) and Software as a Service (SaaS) solutions to a wide range of subsidiary companies. • Led a team of 19 data professionals to create a 35 million entry Iranian E-Commerce dataset. • Spearheaded the creation of the first multimodal search and recommendation system for Persian e-commerce, leveraging AI to understand both text and images.

• Designed and implemented a comprehensive data acquisition and preprocessing pipeline to ensure data quality and consistency for algorithmic trading strategies using RAPIDS.AI stack on GPU. • Developed and backtested quantitative trading strategies using machine learning algorithms, achieving an 8% increase in return on investment. • Optimized the performance of trading algorithms by leveraging Apache Arrow Flight for efficient data transfer between microservices, reducing latency by 30%.

• Developed and deployed a machine learning model to identify and flag suspicious payment patterns, improving the accuracy and efficiency of fraud prevention efforts by 3%. • Developed an ELK based multilingual search engine for e-commerce product listing page that increased overall CTR by 53%. • Collaborated with the product team to design and implement a personalized product recommendation engine for e-commerce platforms, leading to a 31% increase in sales conversion rates.

• Developed and deployed machine learning models to forecast ticket demand, resulting in a 13% improvement in resource allocation. • Collaborated with software engineers to integrate the models into the existing booking platform, ensuring seamless user experience and real-time updates. • Contributed to the development of a data pipeline using Apache Kafka and Airflow to process and analyze large volumes of travel data, enabling efficient monitoring of trends and anomalies.

Thesis: Machine Learning for de novo design of antimicrobial peptides