Role-specific. Annual pricing. Anyone on the team can ping it.

Data Analyst Fellow

A custom-built AI data analyst that lives in your team's Slack. Answers questions about the numbers, writes SQL against your warehouse, builds quick visualizations, and drafts the memo. Trained on your data model, your metric definitions, and your team's analytical style.

What a Data Analyst fellow does

The data questions your team asks 20 times a week, answered.

Q

Answers ad-hoc questions.

Ask about the numbers in Slack, get the answer with the query, the chart, and the source.

S

Writes SQL against your warehouse.

Knows your schema, your dimensions, your common joins. Shows the query so analysts can verify.

V

Builds quick visualizations.

Draws the chart inline in Slack for a Friday-quick read or drops it in your BI tool for permanence.

M

Drafts analysis memos.

Pulls the data, writes the read, flags the caveats. Ready for an analyst to verify.

W

Watches metrics for anomalies.

Pings when a key metric shifts week-over-week beyond your normal noise band.

D

Maintains the metric dictionary.

Keeps definitions current, flags conflicting definitions across teams, drafts updates.

Tools your Data Analyst fellow connects to

Deep, role-specific integrations. Not 3,000 shallow ones.

Warehouse

Snowflake BigQuery Redshift Databricks Postgres

BI Tools

Looker Mode Hex Metabase Tableau (via API)

Modeling

dbt Coalesce

Product Analytics

Mixpanel Amplitude Heap Pendo

Spreadsheets

Google Sheets Excel

Communication

Slack Microsoft Teams Gmail

Docs

Notion Confluence

Need a tool not listed? Tell us during scoping.

How a Data Analyst fellow shows up in your channel

Your first 30 days with a Data Analyst fellow

Week 1

Setup

  • Scope the role with you (warehouse schema, metric definitions, BI tool, analytical style)
  • Custom-train Ash on your data model, dbt project, and prior memos
  • Add Ash to your Slack and connect warehouse + BI in read mode

Week 2

Shadow

  • Ash answers questions in read-only with query + caveats shown
  • You give it small jobs: pull last week's funnel, draft this memo, check this anomaly
  • Every memo reviewed before publishing

Week 3

Delegate

  • Ad-hoc Slack questions answered with the query attached for verification
  • Anomaly monitors run on key metrics and ping the channel
  • Memo drafts auto-generate for standing analyses (weekly funnel, monthly retention)

Week 4

Measure

  • Review questions answered, anomalies surfaced, memos drafted
  • Decide what to expand into Ash's scope
  • Onboard another role if ready

Why this is different from a BI tool with AI chat

Looker and Mode are your system of record for dashboards. Ash is the analyst who answers the questions your dashboards do not cover — the ad-hoc Slack question, the anomaly investigation, the board memo that needs a human read on the numbers.

Most teams run the fellow alongside their BI tool. Ash is the junior analyst you have not hired yet.

Pricing

Each Data Analyst fellow is on a custom annual plan based on data model complexity and team size. Predictable annual price. No per-query meters.

Get pricing for a Data Analyst fellow

Frequently asked questions

Ash always shows the query. Hallucinated numbers are exposed by querying their source. We tune the fellow to flag uncertainty (small sample, ambiguous metric definition, missing data) rather than guess. Verify before publishing externally.

Generic text-to-SQL guesses your schema and your metric definitions. The fellow is trained on YOUR schema, YOUR metric dictionary, and YOUR common joins. Same general approach, very different accuracy.

It can draft a Mode/Hex/Looker dashboard structure and push the queries. Long-term dashboard maintenance is human work — analysts know which dashboards still earn their seat.

It flags them. If 'active user' means one thing in product and another in marketing, the fellow surfaces the conflict and asks which definition you want for this answer.

Bring on a Data Analyst fellow.

Tell us about your data. We will scope the fellow and have it in your Slack in about a week.