Introduction

Datyo is an Agent-native intelligent data analytics platform. Connect your databases or files, ask in natural language, and get executable SQL, tables, charts, and business insights; you can also turn frequent questions into scheduled jobs whose results are pushed on time to Lark, DingTalk, or WeCom groups.

In one line: let your team talk to data directly.

What it solves

• Queries need scheduling: even simple GMV, DAU, or conversion-funnel questions require a request and waiting on the data team. Datyo lets business users ask themselves.
• Monitoring is manual: daily operating reports and key metrics are run by hand. Turn them into scheduled jobs that run on time, with automatic alerts on failure.
• Inconsistent definitions: the same metric is computed differently by different people. Write the definition into a Space and the Agent's answers stay consistent everywhere.
• BI has a learning curve: drag-and-drop modeling and picking chart types take training. Here you just ask in natural language and charts are generated automatically.

What it does

• Connect many data sources: relational databases, warehouses, lake formats, files, object storage, and more, with connection secrets stored encrypted.
• Ask in natural language: translate business questions into SQL and execute them, with multi-turn follow-ups and self-correction.
• Auto visualization: pick the right chart from the question and result, no manual configuration.
• Scheduled monitoring: run batches on any schedule, push results to IM groups, and alert on failure.
• Ask in chat: @mention the bot in a Lark / DingTalk / WeCom group to ask directly.
• Save & share: save results as dashboards and generate a read-only share link in one click.

Who it's for

• Business monitoring / daily reports: push core metrics to a group as scheduled jobs; spot an anomaly and ask a follow-up for minute-level root cause.
• Self-serve data analysis: automate frequent questions so the data team's energy goes to genuinely complex analysis and modeling.
• Eng / product self-serve: check event tracking, error rates, and key metrics yourself, without waiting on the data schedule.

Next steps

• Quick start: run your first analysis in 5 minutes
• Core concepts: understand Spaces, datasets, conversations, and dashboards
• Want to run it in your own environment? See Self-hosting