
Infrastructure,
understood.
Every Windows machine you look after, observed down to the hardware.
A short list of what needs attention.
A controlled, recorded way to act.
Better instruments. Clearer understanding.
Individual signals.
A connected understanding.
Hardware health, security events, remote access and the record of who did what usually live in four places.
Magenta brings them into one console for the machines you are responsible for.
finance-02
Every device sends a heartbeat every 30 s and connects outward to your server; a device that goes quiet is marked offline on its own.

Understanding begins
with better instruments.
We have always built instruments to understand what we cannot observe directly. Magenta brings that same discipline to the systems we depend on.
The thinking behind MagentaYour environment.
Your operational data.
Your terms.
A small agent, Cadence, reports from each machine to a backend and console you host. Every connection starts at the device, every request is signed with a key the device generated itself, and nothing depends on a vendor cloud.
Inside the architecture
Outbound only · Signed · Self-hosted
The work behind the system.
Telemetry for constrained links
Transport for observations that respects thin, unreliable links (our MTS-1 research). It is unpublished and not used by Magenta today: the agent sends JSON over HTTPS.
ExploringState across disconnected systems
Every device already keeps its observations in a local queue until the server confirms them. The open questions are proving delivery across long outages and showing last-known state honestly.
In developmentAssessment that cites its evidence
Rules first, statistical baselines second, models third, with every assessment pointing at the records it rests on. Correlating process, file, user and network activity is a research direction, not a capability.
ExploringA better view starts here.
Request early accessEarly access for IT teams and service providers running Windows fleets.