AI-NATIVE · CERTIFIABLE · AXIOM-POWERED

STERIOD

AI-assisted automation for telemetry data analysis.

Steriod turns plain-language specifications into certifiable, executable Axiom programs — for verification, anomaly detection, root-cause analysis, control, and onboard intelligence.

Get in touch See how it works

Spec → Axiom

AI translates intent into a certifiable rule program.

Multiple use cases

verification, anomaly, root-cause, control, reverse-eng.

Run anywhere

desktop · server · cloud · onboard embedded coming soon.

THE PROBLEM

Complex systems generate data faster than humans can specify, verify, and act on it.

Today's tools force a choice: ad-hoc, expert-locked, or non-certifiable. None gives you all three.

One-off scripts everywhere

Engineers re-write Python or MATLAB for every task — verification, anomaly detection, post-pass analysis. Not reusable, not reviewable.

Specifications drift from reality

Requirements live in Word and Excel. There is no automated path from a written requirement to a running test or controller.

Hours of manual investigation

When telemetry shows an anomaly, engineers grep logs and chase down root cause by hand. Expensive expertise, slow turnaround.

Auditors cannot read the code

Certification depends on artefacts an external auditor can read and sign. Custom scripts and proprietary models do not qualify.

THE SOLUTION

Axiom — the certifiable bridge between human intent and machines.

A domain-specific language for temporal behavioural rules over data streams. Operators describe what the system shall do — Steriod generates a reviewable Axiom program that machines execute.

Human-readable

Engineers and auditors can read a rule with no tool support — it's plain text.

AI-generatable

Generated from natural-language specs, requirements, or learned from telemetry.

Certifiable

Plain-text artefact regulators can review, version, and sign off on.

eps_power_health.axiom

suite is $"EPS Power Health"

expect signal battery_voltage as float

rule $"battery voltage in range" is
    expect battery_voltage >= 12.0 and battery_voltage <= 16.8

rule $"low soc triggers low power" is
    when battery_soc < 20 then eps_mode = 2 within 5 seconds

TRY IT NOW

The Axiom Playground — write, run, and inspect rules in your browser.

No install required. Write Axiom rules, define scripted test scenarios, compile and run them live — then inspect signal timelines and pass/fail results side by side. The fastest way to understand what Axiom can do.

✎

Write rules & tests

Author .axiom suites with temporal rules and scripted let/wait/expect test sequences — syntax-highlighted in the editor.

▶

Compile & run instantly

The Axiom compiler and runtime run entirely in WebAssembly — no server round-trip, no account needed.

◈

Inspect signal timelines

See pass/fail per test, signal values plotted over time, and rule evaluation results — all in one view.

Open the Playground

steriod.ai/playground

Axiom Playground showing code editor with rule definitions and test results alongside signal timeline charts

HOW IT WORKS

One pipeline. Many use cases.

From a written specification to live, certifiable system behaviour — in minutes.

Specification

Plain-language description of what the system shall do, an ECSS clause, or telemetry pattern.

AI agent

Drafts an Axiom program — explains every rule in plain English for human review.

Human review

Engineer edits and approves. Auditor signs the .axiom file as a certifiable artefact.

Steriod engine

Executes the rules against live or recorded data streams. Same engine, every use case.

Outcome + audit

Pass/fail timeline, anomaly flags, control directives, traceable evidence for V&V.

Built on the standards your ground segment already speaks

HOSTS ON

MQTT · Kafka · NATS · HTTP · Files · Custom

ENGINE

Steriod App · Steriod Service Manager · Axiom compiler + runtime (Go today, Rust/C planned)

DEPLOYS

Desktop · on-prem server · cloud · onboard MCU

USE CASES

Start with V&V. Grow without switching tools.

Automated V&V and anomaly detection are available today. Root-cause analysis and diagnostics expand on the same data, the same rules, and the same engine — so your investment compounds as you add capabilities.

AVAILABLE NOW · WHERE TO START

Automated V&V & live anomaly detection

Continuous verification of system behaviour, with real-time flagging of deviations on telemetry streams. One unified rule layer replaces ad-hoc Python and disconnected alarm thresholds.

● Verify rules continuously against recorded or live data streams
● Surface anomalies in real time as rule violations
● AI suggests new rules from observed deviations
● Plain-text certifiable artefact every auditor can read

COMING NEXT

AI-driven diagnostics

Once Axiom is running in your pipeline, the same engine and data unlock:

Root-cause analysis

AI walks the audit trail and proposes the most likely cause in plain English when a rule fails.

Requirement reverse-eng.

Learn rules from historical telemetry to bring legacy systems into spec.

Same engine. More from your data.