Jupyter Notebook Praktikum Projects. This is repository with data analyst educational projects from Yandex.Praktikum.

Financial modeling for AI products. Model LLM inference costs, test pricing scenarios, and see 3-year projections.

This repository is containing a portfolio of data analyst projects that I have completed and showcases my skills and experience

DA test task for Vizor

Works completed within the Product Analytics and the Product Development modules of the Data Analyst course on the Karpov Courses platform

Exploratory data analysis projects

unit economics & cohort analysis projects

Startup metrics for Claude Code — PMF, unit economics, churn, runway, funnels, growth curves, moats, fundraise readiness

Interactive SaaS unit economics calculator with scenario comparison, health scoring, and metric explainers. Built with vanilla JS — no dependencies.

Audit OpenClaw agent spend, waste, and before/after improvements.

Enterprise-grade Power BI dashboard for tracking AI infrastructure unit economics. Features real-time GPU profitability ratios, idle waste detection (Zombie GPUs), and ESG carbon footprint reporting.

Excel financial model analyzing SaaS unit economics, break-even dynamics and growth scalability

Анализ юнит-экономики и бизнес-моделей: CAC, LTV, когортный анализ и оценка прибыльности цифровых продуктов.

Business Studio project: Dodo Pizza market expansion strategy with unit economics, A/B testing, field research. ROI 48-57%, payback 2-3.5 years.

AI Economic Identity Envelope — A standard economic identity for every AI transaction

Cohort CAC/LTV model for GTM analytics. DuckDB SQL executed from Python. Exports Excel + PNG for decision-ready unit economics.

Unit Economics Calculator — это бесплатный, мощный и красивый инструмент для расчета себестоимости, маржинальности и точки безубыточности. Создан специально для мастеров handmade, малого бизнеса, самозанятых и предпринимателей, которые хотят понимать реальную экономику своих продуктов.

High-fidelity digital twin and predictive ROI engine for electric aircraft fleets. Bridges 1Hz physics telemetry with a 96%-accuracy polynomial ML model and linear programming (Solver) to maximize EBITDA under regional energy constraints.

Strategic Decision Engineo optimize profitability for a regional airline facing a 25% fuel price shock. The engine stress tests three strategic horizons Dynamic Pricing, Government Subsidy (UDAN) Expansion, and Fleet Redesign against real-time market volatility to minimize the Risk of Ruin.

Financial planning layer for APPNEURAL budgets, forecasts, unit economics, billing, revenue, and reporting.