AIVAON Connect is a pre-release inbound AI receptionist. Pilot availability is subject to qualification. Live go-live is not authorized in this environment.

How does AIVAON handle phone calls?

When a call arrives via PSTN, SIP, or WebRTC, AIVAON answers in under 1 second. Eight AI agents simultaneously analyze voice for intent, emotion, context, and safety. The AI takes action within the conversation — booking appointments, answering questions, qualifying leads, or routing to a human agent with full context.

Is AIVAON free for consumers?

Yes. Consumer access to AIVAON is always free. Anyone can call or use the help network to get instant AI-powered assistance. AIVAON connects consumers to the right service providers through its Ambient Help Layer — no sign-up required.

What industries does AIVAON serve?

The first bounded pilot is one United States local-service business. Healthcare, legal, financial, insurance, and other regulated decisions are excluded from the first pilot.

Is AIVAON compliant with healthcare and financial regulations?

Controls for HIPAA, PCI-DSS, SOC 2 Type II, and GDPR exist in design. Formal certifications are not issued. Live go-live is not authorized in this environment.

How fast is the AI voice response?

Latency targets exist in design only and are not measured in a named production environment. Live go-live is not authorized in this environment.

What counts as a voice minute?

A voice minute is any minute of active conversation between your AI agent and a caller. Hold time, ringing, and silence are not counted.

Can I bring my own LLM API keys?

Yes. You can use your own OpenAI, Anthropic, Google, or any OpenAI-compatible API keys. Or use our hosted inference at no additional cost.

Is there a free trial?

The Starter plan is free forever with 100 voice minutes / month. No credit card required. You can upgrade to Pro at any time.

How does local-first deployment work?

Enterprise customers can deploy AIVAON on their own infrastructure. AI models run locally, and sensitive data never leaves your network.

What about phone numbers?

Phone numbers are available in 100+ countries via our multi-carrier provisioning. Numbers start at $1/month depending on type and country.

How does pricing work?

Every engagement starts with understanding your challenge. We scope, architect, and price a solution built for your specific operation. For self-serve, AIVAON offers a free Starter tier (100 voice minutes / month), a Pro tier at $0.06/minute with unlimited minutes and 10 ai agents, and custom Enterprise pricing. Consumer access is always free. Partners earn up to 30% revenue share.

Can I deploy AIVAON on my own infrastructure?

Yes. Enterprise customers can deploy AIVAON locally. AI models run on your infrastructure, and sensitive data never crosses network boundaries. This supports air-gapped environments and strict data sovereignty requirements.

What is AIVAON's compliance posture?

AIVAON is engineered to SOC 2 Type II, HIPAA, GDPR, PCI-DSS, and ISO 27001 controls; formal third-party certifications are in progress. Enterprise customers receive a data processing agreement and current compliance documentation.

Release Readiness — AIVAON | Evidence-driven Production Validation

Loading…

The 15 Validation Waves

Validation follows the architectural dependency order (§15.1 consequence two): foundation before control plane, control plane before media, media before AI, AI before telephony, telephony before commercial operation, commercial operation before pilot traffic.

w01

Foundation & Infrastructure

§15.3

w02

Authentication & Tenancy

§15.4

w03

Control Plane & Entitlement

§15.5

w04

Realtime Media

§15.6

w05

Telephony

§15.6

w06

AI Runtime

§15.7

w07

Memory, Workflow, Tool Bus

§15.7

w08

Observability & Billing

§15.7

w09

Offline & On-Device

§15.7

w10

Compliance & Residency

§15.8

w11

Admin & Customer Console

§15.8

w12

Performance, Soak & Chaos

§15.8

w13

Pilot Scenarios

§15.8

w14

Release Gate Review

§15.9

w15

Continuous Validation Post-Launch

§15.10

How the Release Gate Works

Binary, not discretionary. Every criterion is binary — pass, fail, or insufficient_evidence. Release decisions cannot be influenced by schedule pressure.

Evidence-window enforced. The gate refuses to authorize release if any required wave has not produced evidence within the configured window (default 168 hours).

Tamper-evident. Every wave-run summary and gate decision is chained into the §10.6 audit ledger with cryptographic hash linkage; tampering is detectable.

Continuous, not periodic. A synthetic probe runs against every cell every 5 minutes; the gate re-evaluates every 4 hours; the spec-sanity check runs hourly. Validation is a structural property of the platform, not an administrative checklist.

Auditor-grade. Evidence is durable through a Postgres ledger with append-only triggers (UPDATE and DELETE refused at the database level), and evidence bundles are produced on demand for SOC 2, ISO 27001, HITRUST, and regulator inquiry.

For AI Agents & Auditors

The current posture is also available as machine-readable JSON for AI agents and auditor systems:

GET https://www.aivaon.com/api/v1/aivaon/public/release-readiness/posture

The same posture is exposed through the AIVAON Model Context Protocol (MCP) tool aivaon_section15_get_release_readiness_posture for agentic consumers.

The 15 Validation Waves

w01

Foundation & Infrastructure

§15.3

w02

Authentication & Tenancy

§15.4

w03

Control Plane & Entitlement

§15.5

w04

Realtime Media

§15.6

w05

Telephony

§15.6

w06

AI Runtime

§15.7

w07

Memory, Workflow, Tool Bus

§15.7

w08

Observability & Billing

§15.7

w09

Offline & On-Device

§15.7

w10

Compliance & Residency

§15.8

w11

Admin & Customer Console

§15.8

w12

Performance, Soak & Chaos

§15.8

w13

Pilot Scenarios

§15.8

w14

Release Gate Review

§15.9

w15

Continuous Validation Post-Launch

§15.10

How the Release Gate Works

Binary, not discretionary. Every criterion is binary — pass, fail, or insufficient_evidence. Release decisions cannot be influenced by schedule pressure.

Evidence-window enforced. The gate refuses to authorize release if any required wave has not produced evidence within the configured window (default 168 hours).

Tamper-evident. Every wave-run summary and gate decision is chained into the §10.6 audit ledger with cryptographic hash linkage; tampering is detectable.

For AI Agents & Auditors

The current posture is also available as machine-readable JSON for AI agents and auditor systems:

GET https://www.aivaon.com/api/v1/aivaon/public/release-readiness/posture

The same posture is exposed through the AIVAON Model Context Protocol (MCP) tool aivaon_section15_get_release_readiness_posture for agentic consumers.

Release Readiness — Evidence-Driven Production Validation

Current Posture

The 15 Validation Waves

How the Release Gate Works

For AI Agents & Auditors

Release Readiness — Evidence-Driven Production Validation

Current Posture

The 15 Validation Waves

How the Release Gate Works

For AI Agents & Auditors