Murad Aladdinzade

AI Engineer, Capstone Project at SymTrain LLC – IT Services, Nashville, USA

Showcased OpenCV-based computer vision model in AWS SageMaker Notebooks using customer-service training videos. Deployed FastAPI-based AWS Lambda endpoints for model inference, enabling real-time integration across services. Automated detection of UI elements, including buttons and text fields, as well as cursor location, using OpenCV.

AI Engineer (Summer Intern) at Vanderbilt University – Data Science Institute, Nashville, USA

Boosted Llama performance by fine-tuning with PEFT in Python, enhancing response relevance. Built a RAG chatbot using GPT-4 API, LangChain and a FAISS vector database, boosting top-5 retrieval precision by 25%.

Senior Data Scientist at Azerconnect Group LLC – Telecommunications, Baku, Azerbaijan

Developed XGBoost churn model that reduced monthly telco subscriber churn by 1.5%, strengthening retention strategies. Clustered customers with K-means into seven cohorts (2 high-value, 3 mid-tier, 2 low-value), enabling targeted campaigns.

Business Intelligence and Analytics Specialist at Azerconnect Group LLC – Telecommunications, Baku, Azerbaijan

Engineered daily subscriber KPI analytics in R and visualized in Power BI, delivering timely insights to C-level management. Automated generation & emailing of Excel reports weekly via R’s openxlsx, enabling data-driven decision making. Engineered Python ETL pipeline to consolidate daily records and auto-generate KPI reports, saving 20+ hours weekly.

AI Engineer at SymTrain LLC

Shaped AI capabilities with OpenCV-based computer vision model in AWS SageMaker Notebooks; deployed FastAPI-based AWS Lambda endpoints for real-time model inference; automated UI-element and cursor-location detection using OpenCV.

AI Engineer (Summer Intern) at Vanderbilt University – Data Science Institute

Fine-tuned Llama model using PEFT in Python; built a RAG chatbot using GPT-4 API, LangChain, and FAISS, improving top-5 retrieval accuracy by ~25%.

Senior Data Scientist at Azerconnect Group LLC

Developed XGBoost churn model reducing monthly subscriber churn by 1.5%; clustered customers into seven cohorts via K-means for targeted campaigns.