Solantir Infrastructure

The engines underneath everything Solantir ships.

Products are the surface. Underneath Lume and Kestrel sits Solantir's actual technical asset — three infrastructure engines, each responsible for a distinct part of the process of turning raw data into usable intelligence.

Engine 01

Omnis

Solantir's pattern recognition and prediction engine — domain-agnostic by design. Omnis isn't built to understand one specific type of data. It's built to find patterns, detect anomalies, and generate forward-looking insight across whatever structured data it's given, regardless of the industry that data comes from.

That domain-agnosticism is what lets Omnis sit underneath Kestrel today and, in principle, underneath any future product that needs the same core capability applied to a different kind of data entirely.

Active · Powers Kestrel
Domain-agnostic pattern detection

Every dot below is a record. Omnis doesn't care what the records represent — suppliers, transactions, sensor readings — only how they relate. Drag your cursor through it.

LumeX · Live Forecast Model

Detected risk

Mid-term dip

Engine 02

LumeX

The predictive and pattern recognition model that specifically powers Lume. LumeX takes a student's academic data and generates forward-looking insight about where their performance is heading — functioning as the analytical core underneath every forecast Lume shows.

Unlike Omnis, LumeX is purpose-built for one domain: academic performance data. That specialisation is what lets it identify a specific, individual risk — not a generic study tip — and hand it straight to Lume's study-plan generator.

Active · Powers Lume
Engine 03

Spector

Solantir's data construction and orchestration engine — the layer responsible for structuring and organising raw data before it can be meaningfully analysed by Omnis or any downstream system.

Spector is the newest and least developed of the three engines, still in active research, and is intended to eventually power a future product called Sable — built to give order to high-entropy data no matter how complex, vast, or fragmented the source.

In research · Powers Sable (future)
03 / ENGINE

Spector

Structure from entropy

Disordered, disconnected fields resolving into a clean, aligned structure — the same operation Spector performs on raw enterprise data, just visualised.

The Stack

How the engines meet the products.

Omnis →

Kestrel

Supply chain & operational intelligence

LumeX →

Lume

Academic performance forecasting

Spector →

Sable (future)

Order from high-entropy data, at scale

Built at infrastructure depth

Every product starts with an engine, not a feature.

If your organisation has a data problem that looks like Omnis's domain — pattern recognition and prediction across structured data — we want to hear about it.