Proof, from the fragmented to the finished
A representative project showing how we take a hard, multi-source data problem and turn it into a dataset a research team can rely on.
DAO Governance + Historical Market Dataset
We built a research-ready dataset for 400+ DAOs by mapping DAO names to token tickers, extracting historical market data, reconciling missing tokens, testing multiple financial APIs, normalizing prices across USD/BTC/ETH, removing duplicates, and using manual chart extraction where APIs failed.
The workflow behind the dataset
Every step existed to protect one thing: the trustworthiness of the final numbers.
Entity mapping
Mapped 400+ DAO names to their token tickers and contract addresses using fuzzy matching, then reviewed edge cases by hand to eliminate mismatches.
Source testing
Tested 10+ financial and crypto data APIs for coverage, history depth, and reliability — then selected the best source per token rather than trusting a single vendor.
Manual reconstruction
Where APIs had gaps or no coverage at all, we reconstructed historical prices directly from charts and dashboards, and reconciled them against known reference points.
Normalization
Normalized all pricing across USD, BTC, and ETH denominations so tokens could be compared on a consistent basis across the full time range.
Cleaning & dedup
Removed duplicates, resolved conflicting values between sources, standardized the schema, and flagged partial coverage explicitly instead of hiding it.
Delivery
Shipped a research-ready dataset in CSV and Excel, with documentation covering sources, coverage per token, normalization method, and known limitations.
What research-ready
actually looks like.
A sample of delivered structure — normalized, deduplicated, validated. Full datasets are scoped to your specific research question.
// this is a sample preview. full datasets are built or delivered based on client requirements.
Have a dataset that doesn't exist yet?
Tell us the research question and where the data lives. We'll figure out how to build it — book a call or send a request and we'll reply within 24 hours.