Your machine data and business data, in one place.
linalDB is our analytical engine: hybrid tables that mix business columns with vectors and matrices, native signal functions (FFT, PSD, BANDPASS), and a single binary that can run inside your plant.
See the documentation →How we apply it today
This engine isn’t a technical exercise, it’s the foundation of the 5 lines where we sell data analytics:
Why it matters
- Fewer systems to pay for and maintain: business and signal data in the same table
- Vibration analysis happens where the data lives, no exporting to another tool
- Embedded mode: can run on a server or gateway inside your plant
- Persistent lineage: every alert can be traced back to the sensor and computation that produced it
- Community-licensed and inspectable: if we’re not working together tomorrow, your system keeps running
Example query
SELECT id, FFT(vibration_signal) AS spectrum
FROM sensor_readings
WHERE machine_id = 'M-204'
AND ts BETWEEN '2026-09-01' AND '2026-09-22';
Honest scope: linalDB is single-node, built for plants with tens to hundreds of sensors per site. For massive distributed processing, we integrate it with complementary tools.
Frequently asked questions
Is linalDB a product I buy, or part of the service?
It's our own engine: we use it to deliver our analytics services. It's also open source, community-licensed, and free for research and education.
Does linalDB replace my current database?
Not necessarily. It can run embedded, alongside what you already have, joining business data and signal data in a single table.
How much can it scale?
It's single-node, built for tens to hundreds of sensors per site. For massive distributed processing, we integrate it with complementary tools.