Inbound voice
Answer every call with a natural, real-time voice agent — grounded in your knowledge base, never reading from a static script.
Governed Voice AI — Compliance + Brand runtime
What Audactactually does.
A full voice-AI platform — inbound and outbound voice live today, a managed speech stack and knowledge-grounded answers, with chat across web, WhatsApp and email shipping shortly after launch. Plus the proof layer no one else ships: Article 50 disclosure, a cryptographic evidence chain, and real-time brand governance on every utterance. EU-hosted, built for European agencies.
What the platform does
Everything you expect from a modern voice-AI platform —plus the proof layer no one else ships.
Audact is a full conversational platform first: inbound and outbound voice live today, a managed speech stack and knowledge-grounded answers, with chat across web, WhatsApp and email shipping shortly after launch. That live voice product is the foundation. The compliance and evidence layer below is what makes itGoverned.
Outbound voice
Run outbound campaigns with eligibility checks before the first ring, so the call only happens when it is allowed to.
Chat — web, WhatsApp, email
PlannedChat is shipping shortly after launch: the same agent will handle web chat, WhatsApp and email under one set of governance and disclosure rules.
Managed speech stack
Speech-to-text, the language model and text-to-speech are run for you. No API keys to juggle, no vendors to stitch together, no BYOK.
Knowledge-grounded answers
Answers are grounded in your own knowledge base, with per-call transcripts you can read back for every conversation.
Multilingual within the EU
Agents converse in the caller’s European language, with disclosure phrased natively for the jurisdiction they are calling from.
Custom function / tool-calling
LiveYour agent calls your own tools and APIs mid-call; customers add tools themselves.
Retell-compatible webhooks
LiveCall-lifecycle + compliance events, retries and signatures — drop-in for teams on Retell.
In-browser test call
LivePlace a real WebRTC test call from the dashboard before buying a number.
Real-time brand governance runs on every utterance, Article 50 disclosure plays at call-start, and each conversation is signed into a cryptographic evidence chain. Same platform, one proof layer — not a compliance bolt-on.
EU coverage
Voice live in 5 (DE/NL/FR/AT/BE).One Article 50 disclosure runtime across the EU.
Voice agents are live today in Germany, the Netherlands, France, Austria and Belgium. One Article 50 disclosure runtime, with jurisdiction-specific policy configurations across the EU — so when voice rolls out to your country, the disclosure and policy work is already configured.
Voice live today — 5 countries
Article 50 compliance ready, voice rolling out
DEGermanyNLNetherlandsFRFranceATAustriaBEBelgium
IEIrelandLULuxembourgITItalyESSpainPTPortugalDKDenmarkSESwedenFIFinlandPLPolandCZCzechiaSKSlovakiaHUHungaryRORomaniaBGBulgariaGRGreeceHRCroatiaSISloveniaEEEstoniaLVLatviaLTLithuaniaCYCyprusMTMalta
5 live for voice + 22 compliance-ready = all 27 EU member states covered by the Article 50 policy engine on day one.
Article 50 Compliance
Audible disclosure runtime — built into every call.
EU AI Act Article 50 has been in force since 2 August 2026. We treat audible disclosure as a runtime, not a checkbox. Each call opens with disclosure in the caller’s language, recorded against a per-call evidence record, and tied to the policy engine for the caller’s jurisdiction.
Patent-pending (UK IPO) — Article 50 attestation
- Audible disclosure at call-start — never a website notice
- Caller-language detection, multi-language phrasing library
- Per-conversation logged event, signed into the evidence chain
- Per-country policy engine (DE/NL/FR/AT/BE on launch; remaining EU rolling out)
Cryptographic Evidence Chain
Receipt per call. Merkle-aggregated. Exportable on demand.
Every call generates a cryptographic receipt — the prompt, the response, the disclosure event, the compliance score, the policy version. Receipts are Merkle-aggregated into a tamper-evident chain. When a regulator asks, you export the signed proof in under a minute.
Patent-pending (UK IPO)
- Per-call receipt: prompt, response, disclosure event, policy version
- SHA-256 hash chain + Merkle aggregation (patent-pending)
- Exportable as signed PDF + JSON for regulator submission
- PII crypto-shredding aligned to GDPR retention policy
- Pre-dispatch compliance gate prevents non-compliant output
Patent-Pending Portfolio
Depth across compliance, brand governance, and runtime.
Audact filed 44 patent-pending families at the UK IPO between March and April 2026 — covering Article 50 attestation, evidence chain construction, brand-governance state machines, and per-jurisdiction policy enforcement. All are applications under examination; none are granted yet.
44 patent-pending families, filed UK IPO
- Inline proxy, compliance receipt, multi-candidate validation
- Brand governance + commercial transaction validation
- Article 50 attestation + disclosure sequencing
- Per-country regulatory overlay engine
- Follow-on filings on platform governance + risk
Personally Configured Onboarding
Our engineering team configures your first agent.
We onboard each agency personally. Our engineering team configures the agent end-to-end before go-live — disclosure wording, voice, knowledge base, CRM webhook. You operate; we engineer.
Personally configured onboarding for early customers.
- 20-minute scoping call → contract within five business days
- First governed test call the same day from contract signing
- Monthly cryptographic compliance report (PDF + JSON, signed)
- 4 hours/month direct support from the engineering team
- Frankfurt-only data residency, EU-hosted infrastructure
See it in production.
Agency-first GTM. Personally onboarded. Article 50-compliant from day one.