Crypto Analytics: On-Chain Data with Dune, SQL and Python

About this course

**Block Relations Academy — Crypto Analytics: On-Chain Data with Dune, SQL and Python**

Blockchain data is public, but reading it requires skill. Every block, transaction, log, and trace sits out in the open — yet most people still outsource their understanding of on-chain activity to third-party dashboards and APIs that hide the methodology behind the numbers. This course teaches you to query that data yourself: SQL on Dune, Python for deeper analysis, and the judgment to turn raw tables into published research.

You’ll start with blockchain fundamentals — how blocks, transactions, the mempool, logs, events, and traces are structured, and how to navigate Dune’s decoded-table schema instead of guessing at column names. From there you move into SQL built for scale: SELECT and WHERE filtering, JOINs and UNIONs across blockchain tables, GROUP BY with window functions, and CTEs and subqueries that keep complex queries readable instead of unmaintainable.

With SQL in hand, you’ll build inside Dune itself — writing and testing queries in the query editor, assembling multi-chart dashboards, and setting up real-time alerts and refresh schedules so a dashboard keeps working long after you publish it. Two analytical modules put those skills to work on real questions: wallet and cohort analysis (tracking balance changes, grouping wallets by first-transaction date, clustering by behavior) and DEX & stablecoin flow analysis (DEX trade volume and pricing, liquidity pool composition, stablecoin transfers and redemptions, and token holder concentration).

The Python module takes your Dune queries further than the UI allows: exporting results via the Dune API, cleaning and reshaping data with pandas, running statistical tests on cohort differences, and writing up findings as reproducible research — queries included, methodology stated, so someone else can check your work.

The capstone is yours: design an original research question, write optimized queries, build a dashboard, run the statistics, and publish findings with full methodology — graded on query correctness, methodology clarity, visualization quality, reproducibility, and the novelty of the insight.

By the end, you’ll be able to answer your own on-chain questions instead of waiting for someone else’s dashboard to answer them for you — and publish the answer as work other analysts can trust and rebuild. No third-party APIs. No black boxes. Just you, SQL, Python, and the blockchain.

Includes 23 lessons across 7 modules, module quizzes, a graded capstone project, and a shareable Block Relations Academy certificate on completion.