Public profile
@1anda
Machine Learning Engineer | Computer Vision • LLM/RAG
Machine Learning Engineer and KBTU Data Science MSc graduate with hands-on experience in Computer Vision, LLM/RAG, FastAPI, Docker, and CUDA-accelerated ML systems.
Experience, skills, projects, and achievements.
Machine Learning Engineer | Computer Vision • LLM/RAG
Production-oriented RAG system with LangGraph-based query routing, semantic document retrieval, grounded generation with citations, MCP tools, Langfuse observability, DeepEval evaluation, Docker Compose, tests, and CI.
Open projectPersistent memory backend for AI agents built with FastAPI, SQLite/FTS5, hybrid retrieval, token-budget-aware context assembly, Docker Compose, and automated tests.
Open projectCustomer segmentation and recommendation system using K-Means and item-item collaborative filtering. Precision@10 = 0.0141, Recall@10 = 0.1410, Coverage = 0.0630.
Open projectHalyk Bank · 2026
Participated in the Halyk AI Hackathon and received an official certificate of participation.
Open linkKBTU · 2026
Achieved up to 218.27× GPU speedup over CPU while developing large-scale Monte Carlo simulations for European option pricing using CUDA and cuRAND.
KBTU
Jan 2026 — May 2026• Developed a Monte Carlo framework for valuing European options and implemented CPU- and CUDA-based versions of the simulations using Python, CUDA, and cuRAND. • Built a reproducible quantitative pipeline: preprocessing the AAPL option chain, Monte Carlo pricing, Black–Scholes benchmarking, and analysis of results. • Validated Monte Carlo estimates against the Black–Scholes analytical model and evaluated pricing error using MAE and RMSE. • Investigated model accuracy by option type, moneyness, and time
ТОО "Ернамыс-Сакшы"
Jul 2025 — Mar 2026• Developed a computer vision pipeline for RTSP video streams: frame preprocessing, people detection, object tracking, and event analytics based on zones, lines, and motion trajectories. • Fine-tuned YOLO models using data from the company’s cameras: participated in collecting and annotating the dataset, training models under real-world lighting, viewing angles, and motion conditions, and evaluating detection quality. • Collaborated with a backend developer to integrate video analytics results with the database and alerting system for automated
KBTU
2024 — 2026GPA: 3.8/4.0. Focus on Machine Learning, Deep Learning, Computer Vision, NLP, Statistics, and Financial Modeling.
UIB
2020 — 2024Focus on Corporate Finance, Financial Analysis, Risk Management, and Financial Modeling.
