Merey Issenbay

Python Developer Intern

BPMasters Ltd.

Built Django APIs and an administrative panel, and reduced PostgreSQL response time by 25%.

• Designed and developed RESTful APIs using Django, ensuring secure and efficient data handling. • Implemented an administrative panel using Django Rest Framework (DRF) for user data and configurations. • Optimized PostgreSQL queries, reducing response time by 25%.

Backend Developer

Alan Tech

Built backend APIs and automation with Django, FastAPI, and PostgreSQL, and improved data integrity by 25%.

• Implemented error handling and data validation procedures using Python and Django, improving data integrity by 25%. • Developed object-oriented Python scripts to automate system tasks and backend workflows. • Built and maintained scalable RESTful APIs using Django and FastAPI frameworks. • Managed PostgreSQL database, effectively handling high-volume data with optimized queries and indexing strategies.

ML Engineer Intern

Kazakhtelecom

Kazakhtelecom is Kazakhstan's largest telecommunications operator, providing fixed and mobile communications, internet access, and digital services.

Prepared and analyzed network infrastructure data for residential internet planning, including SQL work, maps, and ML dataset preparation.

• Wrote and optimized SQL queries to extract, process, and analyze large-scale network infrastructure data for residential internet planning projects. • Visualized infrastructure and network planning data on interactive maps to support decision-making and improve operational efficiency. • Worked on dataset preparation, cleaning, and feature optimization to improve machine learning model performance and evaluation metrics.

AI Engineer

Mediana

Built an NLP and RAG tutoring assistant, raising answer accuracy from 84% to 92% and cutting median response latency by 40%.

• Designed and built the NLP core of an AI tutoring assistant for school students, implementing a RAG system grounded in the official school curriculum textbooks. • Improved answer accuracy from 84% to 92% on an internal evaluation set of 500 questions. • Optimized the retrieval pipeline through chunking, embeddings, and reranking, cutting median response latency by 40% and improving retrieval hit rate@5 by 25%.

Bachelor, Software Engineering

Astana IT University

MS, Artificial Intelligence

Northeastern University