Контактна інформація

Шукач вказав телефон, ел. пошту та LinkedIn.

Отримати контакти цього кандидата можна на сторінці https://www.work.ua/resumes/15441019/

Досвід роботи

ML/AI Engineer

з 09.2023 по нині (2 роки 6 місяців)
КПІ ім. Ігоря Сікорського, Дистанційно (IT)

I am an Applied Mathematics student passionate about solving complex problems through data, with a primary focus on Data Science. I have hands-on experience building and evaluating machine learning and deep learning models, combining analytical thinking with creative solutions to transform raw data into impactful results. I am eager to continuously grow in modern data technologies.

Key Projects:

-- Chinese ANPR: Built a real-time pipeline for detecting and reading license plates using YOLOv8 and EasyOCR. Parsed and converted CCPD dataset from filename-based bounding boxes into YOLO format. Achieved mAP50–95 = 0.95 on validation and 0.87 on test. Implemented frame-by-frame tracking, text stabilization with deque and majority voting, pre-OCR upscaling, and text normalization. Technologies: Python, YOLOv8, EasyOCR, OpenCV, Pandas, Matplotlib, YouTube-dlp.

-- IMDB Sentiment Classification (BiLSTM + Attention): Built a sentiment classification model with BiLSTM and custom Attention layer. Preprocessed text with cleaning, tokenization, stopword removal, and vocabulary construction. Implemented early stopping, learning rate scheduling, and dropout regularization. Achieved 0.874 test accuracy. Technologies: Python, PyTorch, NLTK, Pandas, Matplotlib.

-- History RAG Chatbot: Developed a RAG system for querying Ukrainian history textbooks with evidence-based answers. Engineered text preprocessing with complex RegEx to clean source texts, optimized retrieval using paraphrase-multilingual-MiniLM-L12-v2, fine-tuned chunking (1200/300) in ChromaDB, and integrated Llama 3.1-8b via Groq LPU for instant inference. Technologies: Python, LangChain, FastAPI, React, ChromaDB, HuggingFace, Groq API.

-- AI Legal Document Constructor (Diia): Built a conversational AI platform automating contract creation. Engineered state-aware backend in FastAPI with Azure OpenAI for context-driven data collection and intent recognition. Implemented hybrid validation using Pydantic and formatters. Designed template ingestion with Groq (Llama 3.1) to parse .docx files and generate user-friendly questions. Technologies: Python, FastAPI, React, SQLite, SQLAlchemy, Azure OpenAI, Groq API.

-- Used Car Price Prediction (Stacking Ensemble): Developed ML pipeline for predicting used car prices. Feature engineering included Regex-based engine specs extraction and brand tiers. Two-stage modeling: CatBoost classifier for outlier probability (AUC 0.8877), then bagged ensemble (LightGBM ×2, CatBoost) with Ridge stacking, achieving validation RMSE 73,225 and Private Score 63159.07806. Technologies: Python, Pandas, Scikit-learn, LightGBM, CatBoost.

-- MyMusicMind: Full-stack web app integrating Spotify, AI, and audio analytics for user music insights. Implemented Spotify OAuth, lyrics extraction (Genius API), Groq LLM topic analysis, PostgreSQL caching. Built analytics: album/playlist audio feature extraction (Essentia/librosa), spider charts, consistency scoring; track clustering (KMeans); art

Освіта

Київський політехнічний інститут імені Ігоря Сікорського

Прикладна математика
Незакінчена вища, з 2023 по 2027 (4 роки)

КПІ ім. Ігоря Сікорського

Факультет програмних систем та прикладної математики (ФПСПМ), Спеціальність 113 «Прикладна математика», Київ
Незакінчена вища, з 2023 по 2027 (4 роки)

Додаткова освіта та сертифікати

Всі курси на Coursera

починаючи з 2023, є сертифікат

Знання і навички

• Python
• PyTorch
• TensorFlow
• Pandas
• NumPy
• Scikit-learn
• SciPy
• Matplotlib
• Seaborn
• Hugging Face
• LangChain
• ChromaDB
• Azure OpenAI
• Groq
• Mathematics
• Statistics
• Algorithms
• OpenAI
• YOLO
• EasyOCR
• SQL
• PostgreSQL
• MySQL
• BigQuery
• SQLite
• Tableau
• Looker Studio
• FastAPI
• Flask
• REST API
• Docker
• Git
• GitHub
• React
• JavaScript
• HTML
• CSS
• Jupyter Notebook
• Transformers
• Team colloboration
• Problem solving
• Self motivation
• Time management
• Continuous Learning

Знання мов

• Українська — вільно
• Англійська — середній

Додаткова інформація

- Graduated from high school with a gold medal for outstanding academic performance.
- INT20H (2025), BEST Kyiv Hackathon -- in the qualification stage, my team ranked 8th out of 44 teams, securing a place in the finals (Top 10). In the final round, we achieved 6th place overall. Developed machine learning models for fraud detection and time series forecasting.

===== Links =====
LinkedIn: [відкрити контакти](див. вище в блоці «контактна інформація»)
GitHub: https://github.com/LatkoArtem
Kaggle: https://www.kaggle.com/latkoartem
Resume: https://drive.google.com/drive/folders/1_nhFYNYRoQPFYWDraKZr4vGA4DC65mU-?usp=sharing