Industrial Intelligence Platforms
Peryx builds the AI layer for industrial operations — turning continuous telemetry into decisions your team can act on. We started with solar because it is the hardest optimisation problem in the physical world.
Who we are
Peryx builds platforms that model physical assets, learn how they actually behave, and turn continuous telemetry into decisions operators can act on.
Not a dashboard. Not a monitoring portal. One architecture, taken deep into a single industry at a time — because a system that does not know the physics of the asset it is watching can only report numbers back at you.
We began with solar. Dense telemetry, silent degradation, weather coupling, mismatched hardware, and money attached to every hour — everything that makes industrial optimisation hard, in one asset class. Solving it there is what makes the rest tractable.
Products
Everything below carries its real status — what is in production, what is being built, and what is designed and scheduled.
Physics-calibrated performance intelligence for solar fleets — every megawatt-hour of loss attributed to a cause.
The modeling core underneath every Peryx platform, being generalised beyond solar assets.
Programmatic access to telemetry, model outputs, and derived intelligence for your own systems.
Grid and distribution asset intelligence for utilities and network operators.
Process and equipment intelligence for asset-heavy production lines.
Capabilities
Ten capabilities that ship together, because none of them is useful on its own.
Models trained on the residual between physical expectation and measured behaviour, so they learn what is anomalous rather than what is average.
A calibrated model of your specific asset, continuously reconciled against what that asset is actually doing.
Every signal ingested at the resolution it was produced, down to individual strings, drives, and sub-assemblies.
Expected versus actual, computed per interval against conditions observed at the asset — not against nameplate.
Degradation identified while it is still gradual, and ranked by the consequence of leaving it alone.
Fleet, site, and component views built for people who have to act on them during a shift, not for a quarterly slide.
Findings routed into the systems your team already works in, with the evidence attached.
Normalised telemetry and model output available programmatically, so intelligence is not trapped in our interface.
Scoped by role and by site, so plant-level detail reaches the people accountable for it and no further.
Where an asset class or failure mode is specific to your operation, the model is built for it rather than approximated.
Technology
Most platforms average telemetry on the way in because it makes storage cheap and dashboards fast. It also destroys the exact signal that identifies a fault. We store what the asset produced and average only at the point of presentation.
A model trained directly on operating history learns your problems along with your normal behaviour. We build the physical expectation first, then let machine learning work on what the physics cannot explain.
An alert that cannot be checked gets ignored after the second false positive. Every finding carries the measurement, the expectation, and the reasoning that connects them, so it can be verified or rejected on the evidence.
Industries
What changes between them is the physics being modeled and the failure modes worth naming — and how far each one has progressed.
Running against real operational data. Deployed and running against real operational data. This is the product you can use today.
Solar in detailWhy Peryx
Both columns describe real systems. The difference is whether anything in the platform knows what the asset was supposed to be doing.
| Dimension | Traditional monitoring | Peryx |
|---|---|---|
| What it tells you | What happened | Why it happened, and what it is costing |
| Baseline | Nameplate ratings or a rolling average of recent behaviour | A physics model of this specific asset under the conditions actually observed |
| Data resolution | Averaged on ingestion to keep storage and dashboards cheap | Retained at the frequency the asset produced it |
| Loss attribution | One aggregate shortfall figure | Separated into named causes and quantified independently |
| Detection | Threshold alarms, which gradual degradation stays underneath | Deviation from expectation, which catches decay while it is still slow |
| Mixed fleets | One portal per vendor, none of which agree on 'normal' | Brands, vintages, and sites normalised into one operating picture |
| Output | A dashboard someone has to remember to open | Ranked findings with evidence, delivered into your existing tools |
What it tells you
Traditional monitoring
What happened
Peryx
Why it happened, and what it is costing
Baseline
Traditional monitoring
Nameplate ratings or a rolling average of recent behaviour
Peryx
A physics model of this specific asset under the conditions actually observed
Data resolution
Traditional monitoring
Averaged on ingestion to keep storage and dashboards cheap
Peryx
Retained at the frequency the asset produced it
Loss attribution
Traditional monitoring
One aggregate shortfall figure
Peryx
Separated into named causes and quantified independently
Detection
Traditional monitoring
Threshold alarms, which gradual degradation stays underneath
Peryx
Deviation from expectation, which catches decay while it is still slow
Mixed fleets
Traditional monitoring
One portal per vendor, none of which agree on 'normal'
Peryx
Brands, vintages, and sites normalised into one operating picture
Output
Traditional monitoring
A dashboard someone has to remember to open
Peryx
Ranked findings with evidence, delivered into your existing tools
Security & trust
Each operator's telemetry, models, and alerts run in an isolated tenant. Nothing is pooled and nothing crosses the boundary.
End to end, from ingestion through storage, using standard TLS and storage-level encryption.
Permissions are scoped by role and by site, so plant-level detail reaches the people accountable for it and no further.
Access and configuration changes are logged, so you can answer who saw what and when.
Multi-tenant cloud by default, with dedicated and in-VPC deployment where residency or policy requires it.
NDAs and standard enterprise data-handling terms are in place before any operational data is shared.
Questions
Including the ones with awkward answers. If something here is missing, ask us directly.
A walkthrough on your own asset class, with a scoped pilot on one to three sites if it's a fit.