Yessil Nurtolgan

@nyessil

Data Analyst interesed in AI-data roles

I am an AI enthusiast and currently data analyst at Kazakhtelecom. On weekends I play board games with friends and single-player games.

Career

Experience: 2 years

Experience

Data Analyst

Kazakhtelecom

Kazakhstan's national telecommunications operator.

Apr 2026 — Present

Builds telecom data dashboards, ETL pipelines, and Python tools for analysis and reconciliation.

• Built real-time monitoring dashboards and ETL pipelines for large-scale telecom subscriber and infrastructure data. • Developed advanced SQL solutions (Impala) with window functions and multi-schema joins across 500K+ records. • Delivered Python tooling for geospatial matching, web scraping, and automated data reconciliation workflows.

Intern - AI Engineer

EdTech Startup

May 2025 — Jul 2025

Developed learning-focused ML models and AI tools, using interaction data to improve model and product effectiveness.

• Developed machine learning models for personalized learning experiences. • Collaborated with educators to integrate AI tools into curriculum design. • Analyzed user interaction data to improve model accuracy and product effectiveness.

Web Developer

Yqylas Films

May 2021 — Sep 2022

Built and maintained responsive websites and supported the team with web-related issues.

• Developed responsive websites using HTML, CSS, and JavaScript. • Collaborated with designers on user-friendly interfaces and visual design. • Maintained and updated websites to ensure performance, reliability, and uptime. • Provided technical support for web-related issues across the team.

Education

MSc, Artificial Intelligence and Adaptive Systems

University of Sussex

2025

BSc, Computer Science

International Information Technology University

2022

Selected projects

Real-Time Infrastructure Monitoring Dashboard

Built a Python/Flask dashboard tracking 50+ servers across five telecom providers, with failure detection under two minutes.

Subscriber ETL Pipeline

Designed an Impala SQL pipeline for 500K+ telecom subscribers, using window functions and multi-schema joins to unify fragmented data sources.

Geospatial Subscriber Record Matching

Created a geospatial clustering algorithm to identify and assign 10K+ missing subscriber records, reaching 95%+ matching accuracy across incomplete address datasets.

Telecom Tariff Web Scraping Pipeline

Developed a multithreaded pipeline processing 500+ URLs per hour, with 98% accuracy on tariff and pricing extraction.

Address Fuzzy-Matching Reconciliation

Automated fuzzy matching of 10K+ addresses against multiple reference databases, reducing manual reconciliation from 40 hours to two hours per cycle.

Adaptive Learning Recommendation Engine

Trained recommendation models on 18K+ learner sessions; a pilot cohort of 2,400 showed a 12% lift in lesson completion. The best offline model achieved 78% accuracy on its top-five recommendations, 21% above the rules-based baseline.

Skills

SQL (Impala)ETL pipelinesPython (Pandas, NumPy)Geospatial analysisKPI reportingMachine learningNLPModel development and evaluationJavaScriptReactHTML/CSSREST APIsGitFlaskWeb scrapingExcelJupyterPower BI / BI dashboardsMultithreading