Practical notes about data analysis, SQL, Python and BI tools - from real projects and follow-up to 2026.
Category roles, the 80/20 rule, tail-cut analysis and demand transference — how to build a defensible assortment matrix instead of relying on buyer memory.
What to track beyond the headline price, rule-based vs. AI repricing, and why matching a rival's flash sale quietly erodes your margin.
Build ABC and ABC-XYZ inventory classification, understand the nine resulting segments, and see what AI tools add to a once-static report.
Why data governance and data quality became the real constraint on AI adoption in 2026 — active metadata, continuous quality monitoring, and a practical way to start.
Which metrics matter, which affordable BI tools fit which budget, and a 30-day plan for small businesses moving from gut feel to real dashboards.
How agentic AI turns pipelines into self-monitoring systems — monitoring agents, orchestration, and how to build a first autonomous workflow.
How Einstein Copilot, Tableau Pulse and VizQL Data Service transform BI workflows — natural language queries, proactive insights and API access to your data.
From raw CSV to a shareable web app in under an hour — charts, filters, KPI cards and deployment to Streamlit Cloud without any front-end skills.
Build RAG chatbots over your CSV files, connect natural language to SQL and automate report summaries — all in a visual drag-and-drop interface.
Complete A/B testing workflow: hypothesis formulation, sample size calculation, z-test in Python, p-value interpretation and common pitfalls to avoid.
ROW_NUMBER, RANK, LAG, LEAD, running totals and moving averages — every window function an analyst needs with real use cases and ready code.
Pricing, features and learning curve compared — practical guide to choosing the right BI tool for your workflow in 2026.
Build a retention matrix, heatmap and churn calculation from raw transaction data — complete guide with pandas code.
Modern SQL transformation with dbt Core: models, tests, documentation and lineage — no engineering background needed.
How to use ChatGPT, GitHub Copilot and PandasAI to work 3× faster — with real prompts and code examples.