What buyers can trust today
InfraVero is evolving, so this page separates shipped posture from roadmap ambition. It is meant to answer security, deployment, data, and support questions before a deeper technical review.
Deployment, custody, AI, and controls
A concise public view of the operating promises InfraVero can make based on what is shipped today.
On-prem deployment model
InfraVero is built to run close to the infrastructure it monitors, on an on-site appliance NTALabs supplies and maintains. The default story is local ingestion, local storage, local dashboards, and optional secure remote access through approved tunnels.
Data retention
Current shipped defaults retain sensor records for 730 days across recent and long-term storage. Video artifacts are shorter-lived by default and operator-managed per deployment.
AI and privacy
InfraVero Intelligence runs inside your deployment. Answers cite the readings behind them, and connecting an external AI provider is an explicit deployment decision, never a hidden default. Your data is not used to train it.
Supported devices
Shipped integration language is intentionally narrow: sensor telemetry and camera streams today. Broader industrial integrations remain roadmap or pilot-specific until implemented.
Security controls
The current public control set: authentication, MFA, role permissions, encrypted secrets, isolation between workloads, and an outbound-only network boundary. See security & deployment for how each one works.
Backup and restore
Database backup scripting exists today, but full automated disaster recovery across every storage layer and volume should be validated as part of a production readiness review.
Shipped integration matrix
The matrix is intentionally conservative. Anything outside this list should be scoped as a pilot integration or roadmap item.
What's next
Real, tested code that ships behind a deployment flag and rolls out per deployment. Described in the future tense on purpose: it is not part of a default install yet.
More independent checks
Additional checks join the consensus, including ones that recognise a bad pattern by its shape without being trained on your specific equipment.
Detection that adapts
Alerting that continuously tunes to each asset’s own normal, so thresholds fit your equipment rather than a textbook baseline.
Forecasting with confidence bands
Multi-horizon forecasts with a calibrated confidence range, so “probably fine” and “almost certainly a problem” never look the same on screen.
Controls worth validating
Production reviews should validate credential rotation, internal API keys, public endpoint exposure, audit trails, and service readiness before the deployment is used for customer operations.
Security and deploymentPublic guardrails
- No public claim of SOC 2, HIPAA, or similar certification until an actual audit is complete.
- No claim of broad industrial protocol support until those adapters are shipped.
- No promise of one-click disaster recovery until backup coverage includes every stateful tier.
- No promise of dynamic retention or switching the AI provider where today changes are operator-managed.
- Forecasting and self-tuning detection are described as “coming”, not live, until they run on a default install.
Need a deployment review?
Bring the trust center into a pilot call and we will map which controls, sizing assumptions, and backup gaps need validation for your environment.