Data / events over time
Timeseries Database
Turn everyday activity into insights. A Timeseries Database records events with timestamps so you can see not just what happened, but how it changes over time.
The idea
Imagine running a small online business. You know how many orders came in today—but are sales improving? Which days are busiest? Are customers spending more or less? A timeline of daily sales, visitors, and orders turns those separate readings into a pattern you can act on.
Business Pulse
A snapshot of your shop, one day at a time.
Sales revenue · 30-day sample period
$17,103
28% across this period (sample data)
Daily sales
Try it: switch metrics, change the date range, and hover a point to inspect one day. Add a sale to update today's sales and order readings.
1 · Capture
Order completed
Record the value and when it happened.
2 · Organize
Timestamp + readings
Each point joins a timeline.
3 · Understand
See the trend
Compare days, weeks, or months.
Behind the scenes
Every event gets a timestamp.
A sale is stored with its value and the exact moment it happened.
Then
Readings form a timeline.
Query a time range to see daily totals and compare periods.
So
Patterns become visible.
Spot changes that a single snapshot would miss.
Good for
- Usage and billing meters
- Application observability
- IoT and device readings
Implementation notes
- Write immutable points with a timestamp and dimensions.
- Set retention rules for high-volume streams.
- Aggregate before displaying charts or reports.
Implementing it
With a coding agent
Ask your agent to track activity over time
Copy this prompt, fill in the bracketed details, and paste it into your coding agent. It describes the goal, what to record, and the guardrails.
Bracketed text marks details only you can fill in. 797 / 6,000 characters.
Good foundation · 68/100
Traditional SDK / API
Use it from a function
await ultrahost.timeseries.write("llm_tokens", {
value: 842,
dimensions: { model: "reasoning", route: "/api/chat" },
timestamp: new Date(),
})Key takeaway
A Timeseries Database doesn’t just tell you what happened. It helps you understand when it happened and how things are changing.
Next primitive
Cron