Aranthos
Aranthos provides independent, external, and deterministic behavioral measurement and forensic evidence for autonomous systems in production. The platform continuously measures system behavior against its own learned baseline, computing a Drift Health Index (DHI) that quantifies departures from normal operation. Evidence packets are sealed with content signatures and independent timestamps, verifiable by anyone without an Aranthos account.
- Data & Analytics
- Regulatory & Compliance Technology
- Cybersecurity
- Software Only
Funding
Funding not disclosed
across 1 round
Founders
Founder details are not available yet.
Product
Problem
Autonomous systems in production can quietly drift from their own learned baseline behavior while infrastructure dashboards remain green, creating risk for organizations. When incidents occur, teams lack trustworthy, tamper-evident records of what the system actually did, making it difficult to prove what happened — or that nothing did — to auditors, regulators, or courts, even years after the event.
Solution
Aranthos provides an independent, external, and deterministic behavioral measurement and forensic evidence platform for autonomous systems. The platform reads system actions, tools, scopes, outputs, and timing — never prompts, weights, or internal state — and continuously measures behavior against the system's own learned baseline, computing a Drift Health Index (DHI) that quantifies the distance between current and normal behavior. The measurement loop uses zero AI; DHI is a computation that anyone can recompute from the sealed record and arrive at the same score. Findings route directly to existing team workflows with details on which system, which dimension departed, since when, and against which baseline. The platform seals evidence packets with content signatures and independent timestamps from three clocks, enabling tamper-evident verification that does not depend on Aranthos's survival.
Target Audience
Primary customers are organizations operating autonomous systems in production who need independent verification of system behavior for risk management, audit, and legal evidence — including compliance teams, engineering leaders, and legal counsel responsible for demonstrating what autonomous systems did or did not do.
Features
- Drift Health Index (DHI)
- a single number per system (0 to 100, high is healthy) computed as a deterministic measure of distance from the system's own learned baseline, with no AI or model grading in the measurement loop
- Read-only external instrumentation that observes actions, tools, scopes, outputs, and timing, with an explicit guarantee to never read the prompt, weights, or internal state
- Evidence Packet (.EVP) verification with four checks
- seal, signature, timestamp, and DHI recomputation
- verifiable using only the file, with no account required
- Signed gap records that document when a source went dark, with bounded start-to-end times and signature protection
- Three independent timestamp clocks (declared, received, attested) that are never merged, providing independent attestation of when measurements occurred
- Aranthos Connect integration via OpenTelemetry, the open standard already emitted in production, with full-fidelity measurement
- nothing sampled, nothing silently dropped
- Zero AI principle ensures that any auditor can recompute any DHI value from the sealed record and land on the same score