Built for real conversations
Four capabilities that make Veqa feel like a real conversation — not an IVR menu reading a script.
Natively multilingual
English, Spanish, French, and Cantonese without language-switching menus. Language is detected from the first utterance and routed automatically.
Interruptible with grace
Customers can cut in mid-sentence. The agent stops talking within 30–60 ms, listens, responds, and resumes naturally.
Full-call task memory
Multi-step procedures are tracked as structured task plans. On resume, the agent picks up at the exact step it left off.
Emotion-aware speech
TTS intonation adapts to message sentiment: confirmations of delays sound apologetic; good news sounds genuinely warm.
Multilingual at launch
Four languages today, with the architecture to add more without retraining the orchestration layer.
English
- ASR
- NVIDIA Parakeet TDT (streaming)
- TTS
- F5-TTS
Spanish
- ASR
- faster-whisper Large-v3 Turbo
- TTS
- Kokoro
French
- ASR
- faster-whisper Large-v3 Turbo
- TTS
- Kokoro
Cantonese
- ASR
- SenseVoice (FunASR)
- TTS
- CosyVoice 2
Deployed where your data needs to be
Two subscription models. Machplace owns the hardware in both. No third-party Voice AI APIs are ever in the call path.
Veqa Edge
On-Premises Subscription
A turnkey appliance pre-integrated with NVIDIA RTX PRO 6000 Blackwell GPUs, deployed at your facility under a monthly subscription. Machplace retains hardware ownership throughout the lifecycle — you never procure a GPU.
- Voice audio and transcripts never leave your network
- Machplace owns and refreshes the GPU appliance throughout its lifecycle
- Air-gapped configurations available for sensitive workloads
- Architecture compatible with HIPAA, SOC 2, PCI DSS, attorney-client privilege
- Delivered as a containerized stack with documented runbooks
Veqa Cloud
Multi-Tenant Managed Service
The same Veqa service, operated by Machplace on dedicated U.S.-based NVIDIA hardware. Multi-tenant by design — GPU capacity is pooled across customers — but every customer runs inside its own isolated container, and call audio is held only in volatile memory for the duration of the call.
- Per-customer container isolation (Linux namespaces, cgroups, NVIDIA Container Toolkit)
- Process- and namespace-level isolation per customer on pooled GPU capacity
- GPU working buffers cleared between calls
- Zero persistence — voice and transcripts held only in volatile memory, never written to disk
- TLS 1.3 end-to-end + mTLS internal mesh
Industry-agnostic by design
Veqa adapts to your domain via your existing knowledge base — no custom model training required to launch.
Bring Veqa to your call flow.
We're onboarding a small cohort of early-access partners in healthcare, financial services, and regulated B2B. Tell us about your call volumes and compliance requirements and we'll be in touch within one business day.
[email protected] · St. Petersburg, Florida