Last updated August 29, 2026
The datasets here are free. Download them, analyze them, use them for school projects, portfolios, random ideas, whatever you wish. They were built by programmatically scraping public sources, so they can contain errors. Insights published by the community are opinions, not verified facts, even when a certified reviewer has attested to them. Nothing on this site should be treated as professional advice, and you are responsible for how you use it.
CodeSights is a data science platform with two parts: a catalog of free datasets, and a community that shares short insights about them. The primary intent of the platform is for analysts, data scientists, statisticians and the curious to explore the data, surface something interesting, and use their published insights as a way of showcasing their skills. It is a place to practice and demonstrate the craft of analysis, not a source of record.
Every dataset is built by programmatically scraping and structuring publicly available sources (each dataset lists its sources and the date it was last pulled). Automated scraping and parsing leave room for error: values can be misread, fields can be incomplete, and the underlying sources themselves change over time and contain their own mistakes. Datasets are versioned so you can see when data was pulled, but no version is guaranteed to be accurate, complete, or current. If a number matters to you, verify it against the original source.
The datasets are completely free. Downloading requires a free account; once downloaded, you may use the data however you wish: coursework, personal projects, portfolios, experiments, publications, anything. Attribution to CodeSights is appreciated but not required. Note that the underlying sources (for example, Wikipedia) may have their own licensing terms that apply to the source material itself; respecting those is your responsibility.
Insights are short, free-form observations written by community members based on their own analysis. They are published as submitted. Upvotes reflect community interest, not correctness. An insight being popular, well-written, or attached to a well-known dataset does not make it true.
Published insights cannot be edited, by anyone, ever. What you claimed, and when you claimed it, stays as written; that is what makes an insight, its upvotes, and its certified reviews credible. If you no longer stand behind an insight, you can delete it at any time, which also removes its upvotes and attestations. Deletion is permanent.
Anyone can apply to become a certified reviewer. Accounts may hold one pending application at a time; submissions are reviewed personally and decisions are at our discretion. After a decision, an account must wait 30 days before applying again.
Some insights carry an attestation from a certified reviewer, a community member whose background we have checked and who states that they independently reviewed the analysis. An attestation is a signal of peer review and nothing more. It does not make an insight accurate, complete, or reliable, and it must not be held as truth. We are not responsible for the use of any insight, attested or otherwise.
The creator and operator of this site accepts no liability, to the fullest extent permitted by law, for:
Everything on this site is provided as-is, without warranty of any kind. If you rely on it, you do so at your own risk.
Submit your own analysis, keep insights honest about what the data does and does not show, and link your work when you can. Three rules apply to every insight:
The insights board is moderated. We reserve the right to remove any insight that we find to violate these rules, and to remove content or ban accounts for spam, plagiarism, foul language, or deliberately misleading insights. Removal decisions are at our sole discretion, and we do not owe an explanation for them.
CodeSights does not promote pointing AI at a dataset and publishing whatever comes back. The value of this community, and of your insights as a showcase of your skills, comes from your own curiosity: form your own hypotheses, ask your own questions of the data, and follow the threads that genuinely interest you. AI is welcome as a tool for coding your way to the answer, writing the queries, the scripts, the charts, but the creativity that gets you to the insight should be yours. An insight generated blindly by AI showcases nothing about you.
These terms may be updated as the platform evolves; the date at the top reflects the latest revision. Continuing to use the site after a change means you accept the updated terms.