06 Cloud dashboard

Starting point

The four Module 05 queries run in the ClickHouse Cloud SQL console. Your Cloud role can save queries and create dashboards.

Goal

One dashboard named Polymarket market pulse with four current views and a one-minute line chart. No local dashboard product is required.

Step 1 — Save the four investigation queries

In the SQL console, run each Module 05 block and use Save with these exact names:

1. Current probability
2. Five-minute movers
3. Spread and freshness
4. Volume velocity

Do not change the SQL while saving; the instructor uses the names to verify the room.

Step 2 — Save the one-minute chart query

SELECT
    c.minute,
    m.token_id,
    m.question,
    m.outcome,
    round(argMaxMerge(c.close) * 100, 2) AS probability_percent
FROM polymarket.market_midpoints_1m AS c
INNER JOIN
(
    SELECT token_id, question, outcome
    FROM polymarket.markets FINAL
) AS m ON m.token_id = c.token_id
WHERE c.minute >= now() - INTERVAL 30 MINUTE
GROUP BY c.minute, m.token_id, m.question, m.outcome
ORDER BY c.minute, m.token_id;

Save it as One-minute probability and choose a line visualization. Use minute as the time axis, probability_percent as the value, and question/outcome/token ID as the series so identically worded markets stay separate.

Step 3 — Assemble the dashboard

1. Open Dashboards in the ClickHouse Cloud console.
2. Create Polymarket market pulse.
3. Add Current probability as a table or number view.
4. Add Five-minute movers as a table sorted by absolute movement.
5. Add Spread and freshness as a table.
6. Add Volume velocity as a table or bar chart.
7. Add One-minute probability as the line chart.
8. Set the dashboard refresh to the shortest option available in your console.

Step 4 — Verify objective evidence

The dashboard is complete when:

• at least one watched market appears;
• probability, bid, and ask values are not null;
• the freshness age is visible; and
• the line chart advances after the next live or fixture event.

If Dashboards is unavailable

Keep the five saved queries and their SQL Console visualizations. Run this completion check:

SELECT
    count() AS watched_outcomes,
    max(event_at) AS newest_tick,
    dateDiff('second', newest_tick, now()) AS age_seconds
FROM polymarket.price_ticks
WHERE midpoint > 0;

Completion fallback: watched_outcomes > 0 and age_seconds < 120 in fixture mode or an active live feed. This is the only fallback; do not install a local dashboard.

Done when

The dashboard has five views, or all five saved SQL Console visualizations exist and the completion query is fresh.

Next: put an AI agent on the live data.