Lantide Data
← Back to blog
Integrate with Lantide

Why Do You Still Need Lantide When Using Notion?

Notion is suited to PRDs, knowledge bases, and database charts, but it is not a local SQL engine. Run rerunnable analysis in Lantide, then put the HTML or Markdown result back in Notion so planning pages cite evidence-backed conclusions.

Notion is a good home for PRDs, project plans, team knowledge bases, database charts, API sync, and embedded dashboards. See Chart view and Developer Platform.

Its strength is collaboration and documentation. When a page's “18.5% conversion rate” needs to be rerun, traced to SQL, and accompanied by limitations, run the analysis in Lantide and put the result back in Notion.

Why numbers on planning pages become contentious

1. The conclusion and the extraction are separated

After writing “June paid conversion was 18.5%” in Notion, discussion stays in page comments while the actual SQL lives elsewhere. When the data or definition changes, the document does not automatically know which query to rerun.

2. Charts serve Notion databases, not arbitrary analytical warehouses

Charts are useful for task status and content properties, with limits such as grouping counts; they are not an engine for window functions and cross-source joins over millions of events. See Chart view.

3. Comments are not analysis approval

An inline comment can discuss whether to exclude test accounts, but it does not produce a Plan or Report with evidence. See Why formal analysis starts with a Plan.

How to combine them: Notion for context, Lantide for execution

There is currently no native two-way sync. A common workflow is:

  1. Organize the question and background in Notion.
  2. Create the Plan, annotate it, and execute in Lantide to produce a Report / HTML.
  3. Paste the Markdown summary or attach standalone HTML in Notion.
  4. When source data changes, run Source Run in Lantide, then update the Notion attachment or section.
Notion is good at Lantide is good at
PRDs, discussion, tasks, and knowledge bases Reviewable SQL over local CSV / warehouse data
Database properties and charts Plan approval, SQL evidence, and Source Run
Embedded BI / product analytics charts Analysis execution and delivery contracts

Plan / Report and annotations

Lantide uses Markdown for Plan.md and Report.md. In Visual mode, highlight a definition and annotate it; the Agent revises the annotation, then you review it and press Approve & Execute. An annotation targets a specific change and does not guarantee that every other paragraph remains untouched. See USER_GUIDE §11.

HTML / Markdown export

Export a standalone HTML report, including charts and related assets, or copy Markdown into Notion's file block or page body. Notion remains the archive and collaboration surface; live extraction and reruns stay in Lantide.

A practical three-step workflow

  1. Think in Notion: write the decision question, hypothesis, and stakeholders; mark unverified numbers as hypotheses.
  2. Execute in Lantide: create a project and Plan, align denominator and time window, approve, execute, and produce a Report with SQL evidence. See Six sections of a trustworthy AI report.
  3. Archive in Notion: paste the summary, attach HTML / Markdown, and state the data period and rerun method. When it changes, rerun in Lantide and replace the attachment.
Dimension Notion primarily With Lantide
Planning and discussion Yes —
Database / project charts Yes —
Local SQL and lineage — Yes
Formal analysis evidence Screenshot or plugin Plan / Report / Source Run
Handoff API / Workers / manual Export and paste back

Conclusion

Let Notion remain the team's knowledge and planning hub; let Lantide be the execution and verification environment for data claims. A number quoted on a page should point to a rerunnable analysis, not a sentence no one can reconstruct.

Next steps

  • Read A trustworthy AI analysis report.
  • Before publishing the next data conclusion in Notion, complete Plan → Report in Lantide and then paste it into the page.

References