Projects
Selected work.
Lakebase · OLTP · Reverse ETL
Lakebase POC Accelerator
A Databricks Application that streamlines testing and deployment of customer OLTP workloads to Lakebase — Databricks' managed Postgres — with a focus on reverse ETL use cases.
View on GitHub →Performance Engineering · Benchmarking
Metimur — Databricks SQL Benchmark Accelerator
A Python-based accelerator that streamlines data generation and query performance evaluation across Databricks SQL warehouse types. Automated workflows, TPC-DS/TPC-H support, and an AI/BI metrics dashboard.
View on GitHub →Spark ML · Prediction
Predict Churn of a Music Hosting Service
Analyzed customer activity for a music service using Spark DataFrame and Spark SQL on a 247.6 MB dataset in IBM Watson Studio. Built an ML pipeline with Spark ML reaching 0.74 F1 and 0.8 accuracy.
View on GitHub →Deep Learning · Classification
Flower Classification with PyTorch
A neural network trained on ~6.5k images across 102 flower species using torchvision pre-trained models (VGG16, VGG19, DenseNet121), reaching 89% accuracy on the test set.
View on GitHub →Deep Learning · Prediction
Stock Price Prediction
Time-series deep learning models — RNN, LSTM, and GRU — built for stock price prediction.
View on GitHub →