Wordcab

What Wordcab does

Wordcab is an ASR-native private voice AI company founded around 2020 by Aleks Smechov (CEO) and Federico Nitidi (COO), operating from New York with an 11-50 person team. The company packages a four-layer voice AI stack — Wordcab Voice (STT/TTS runtime), Wordcab Think (LLM inference), Wordcab Adapt (fine-tuning and evaluation), and Wordcab Redact (PII/PHI/PCI redaction) — into a licensed container stack that customers self-host inside their own VPC, on-prem Kubernetes, airgap, or hybrid environments. The stack is built on open ASR and LLM models (Qwen3-ASR, Voxtral Realtime, Cohere Transcribe, Qwen3.5, Gemma 4, DeepSeek V3.2, Llama 3.3), supports multiple serving backends (vLLM, SGLang, NVIDIA Triton, ONNX Runtime, SambaNova RDU), and exposes an OpenAI-compatible API surface so customers can migrate by changing two lines of code. Compliance is designed in: SOC 2 Type II audit path, GDPR, and HIPAA BAA are positioned as deployment-ready rather than retrofitted.

Wordcab sells primarily through enterprise field sales into regulated verticals — healthcare, banking and financial services, contact centers, and sales/revenue intelligence — with an additional secondary segment for inference providers seeking white-label voice infrastructure. The go-to-market is anchored by a "Talk to an Engineer" technical entry point, an 8-week typical pilot-to-production timeline, and a channel ecosystem of voice-agent frameworks (Pipecat, LiveKit, Vapi, Retell) and CCaaS platforms (Genesys, Five9, Zoom Phone, Twilio). Revenue is generated through subscription licensing of the container stack, recurring Cloud API subscriptions (Pilot and Production tiers), and attached professional services and Gold Support. ARR was reported at $660K with a $2M valuation in late 2025, consistent with an early-stage infrastructure company still investing in product depth and regulated-vertical pipeline.

The company is backed by LAUNCH (Jason Calacanis as lead investor), LAUNCH Syndicate, Calm Company Fund, and strategic angels from voice, ML infrastructure, CCaaS, and regulated SaaS. Notable customers and partners include Bigtincan, Jiminny, Workiz, SCX.ai (sovereign AI in Australia on SambaNova RDU), Sardius Media, Enthu.ai, and Knowledgator (co-developer of the GLiNER-PII redaction model behind Wordcab Redact).

Name

Wordcab

Legal name

Wordcab

Website

https://wordcab.com

Company type

Private

Founded year

2021

Operating status

Operating

Headcount range

11–50 employees

Short description

Wordcab packages private voice AI infrastructure — ASR, LLM, fine-tuning, and PII redaction — as self-hosted container stacks for healthcare, banking, contact center, and revenue intelligence teams that need customer-controlled deployment.

Ownership category

akta.pro rank

Product category

Voice AI Infrastructure

NAICS

Software Publishers (5132), Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (5182)

SIC

Services-Prepackaged Software (7372), Services-Computer Programming Services (7371)

akta.pro primary industry

Conversational AI Platforms (chat/voice bots, orchestration) (HDAAAFAC)

akta.pro secondary industries

Audio & Speech Analytics (call analytics, QA, insights) (HDAAAFAH), Cloud PBX & Hosted Telephony (HDAIAHAB), Voice Cloning, Dubbing & Audio Generative Media (HDAAAFAK)

Keywords

• Voice AI infrastructure
• Speech-to-text services
• Private voice deployment
• LLM inference platform
• PII redaction software

Headquarters

HQ city

New York

HQ country

United States

HQ region

North America

Markets served

GTM type

B2B

Offering type

Software

Cost components

Technology or R&D, Personnel, Infrastructure, Marketing or Sales, Operations

Revenue model

1. Licensed Container Stack: Wordcab ships as a licensed container stack for self-hosted deployment in customer-controlled infrastructure. Revenue is generated through software licensing of the deployment package (Helm chart, operator, model pools). Gold Support and custom engineering work are packaged on top of the product license.
2. Cloud API (Pilot Tier): Cloud-hosted API access for evaluation and pilot workloads. Pilot tier with rate limits of 60 requests/min, 100,000 tokens/min, and 5 concurrent streams. Provides pathway to Production tier (1,000 requests/min, 1,500,000 tokens/min, 250 concurrent streams).
3. Professional Services and Custom Engineering: Custom engineering work packaged alongside the core product — deployment hardening, security review support, integration assistance. Not a substitute for the product but attached to it.

Pricing tiers

Go-to-market motion1 record

Distribution channels5 records

Marketing channels5 records

Core offering

Wordcab packages a private voice AI stack as licensed production-ready containers for deployment inside customer-controlled infrastructure (VPC, on-prem, airgap, or hybrid). The offering combines four product layers: Wordcab Voice for streaming and batch speech-to-text, text-to-speech, and diarization; Wordcab Think for open-model LLM inference and reasoning; Wordcab Adapt for fine-tuning and evaluation on real customer audio; and Wordcab Redact for PII, PHI, and PCI redaction. The stack is OpenAI-compatible at the API surface, supports multiple open ASR and LLM models, and is sold to regulated enterprises that cannot route voice data through hosted APIs.

Product overview

Wordcab is an ASR-native team packaging private voice AI infrastructure for regulated enterprises. The product architecture is a platform-plus-modules model: Wordcab Platform serves as the unified control plane orchestrating deployment, operations, and routing across four core modules. Wordcab Voice handles the speech runtime (STT/TTS/diarization). Wordcab Think provides the LLM inference layer (chat completions, embeddings, reasoning). Wordcab Adapt enables fine-tuning and evaluation on customer audio. Wordcab Redact handles PII/PHI/PCI detection and masking. All components deploy as licensed container stacks into customer-controlled environments (VPC, on-prem Kubernetes, airgap, hybrid) — no audio or transcripts leave the customer boundary. The stack is OpenAI-compatible at the API surface, supports Python and TypeScript SDKs, and ships with Helm packaging, Terraform modules, and observability tooling for platform engineering teams.

Differentiator

Problem solved

Functional benefit

Products and services

• Wordcab Platform

Quantifiable outcome

• 8-week pilot-to-production timeline vs. 6+ months building in-house
• +4 more outcomes

Named customers5 records

Segments5 records

Ideal customer profiles5 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration22 records

AI capability9 records

Feature6 records

Partnerships

Nine partnerships are on record, tiered flagship, core and major.

• SCX.aiflagshipStrategic or Co-development Partner · 1 January 2026Strategic co-deployment partner for sovereign AI infrastructure. Wordcab Voice deployed on SCX.ai's SambaNova RDU-backed infrastructure in Australia. Joint deployment delivers private voice AI for Australian financial-services and public-sector customers. Sub-100ms ASR latency achieved. Custom models tuned for Australian accents and financial-services vocabulary. Public provider surface at sam.scx.ai.
• KnowledgatorcoreStrategic or Co-development PartnerCo-authored the GLiNER-PII model family (Opir.ai guardrails model family release) with Wordcab. Knowledgator collaborated on the open GLiNER model family for PII, PHI, and PCI detection. The redaction model is co-developed and deployable inside the customer's own boundary.
• PipecatcoreTechnology or IntegrationNative integration. Wordcab Voice and Think ship as first-class Pipecat services. Tested against Pipecat 1.0 (April 2026). Full async pipeline support with VAD, turn detection, and interruption handling. Pipecat's Daily transport also supported.
• LiveKitcoreTechnology or IntegrationNative drop-in provider plugin for LiveKit Agents. Works with LiveKit Cloud or a self-hosted SFU. Audio stays inside the LiveKit instance with no round-trip to a hosted STT vendor.
• Genesys CloudmajorTechnology or IntegrationAudioHook protocol adapter streams Genesys call audio into Wordcab Voice for real-time transcription, QA, and redaction. Supported integration.
• Five9majorTechnology or IntegrationVoiceStream API adapter. Transcripts and QA signals delivered via webhook or Kafka back to Five9 reporting.
• NVIDIA (Triton)majorTechnology or IntegrationNVIDIA Triton supported as a model serving backend. TensorRT-LLM and ONNX Runtime backends for teams standardized on Triton. Model repository layout and config files included in the Helm chart.
• SambaNova SystemsmajorTechnology or IntegrationSambaNova RDU (Reconfigurable Dataflow Architecture) supported as deployment target via SCX.ai partnership. Sub-100ms ASR latency and 60-80% lower power-per-inference vs. equivalent GPU reported on voice workloads.
• TwiliomajorTechnology or IntegrationNative integration via WebSocket media streams. Bridge Twilio Media Streams (μ-law/8kHz) into Wordcab Voice. Reference app ships with full TwiML setup and Pipecat-compatible transport.

Scale indicators6 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• Deepgram: Closest direct competitor: cloud and self-hosted ASR with a developer-focused API, model fine-tuning, and enterprise deployment options targeting the same speech-AI-for-engineers buyer as Wordcab.
• AssemblyAI: Speech-to-text and audio intelligence API used by developers and enterprises for transcription, summarization and content moderation; overlaps with Wordcab's STT and LLM summarization layers.
• Speechmatics: ASR vendor with strong on-premises and private-cloud deployment options for regulated industries; directly competes with Wordcab on private-deployment ASR for healthcare, legal and government buyers.
• NVIDIA Riva: GPU-accelerated speech AI SDK for on-prem and private-cloud deployment; comparable to Wordcab's self-hosted ASR/runtime stack on NVIDIA Triton and comparable enterprise infrastructure.
• Observe.AI: Conversation-intelligence platform for contact centers built on ASR, summarization, QA and redaction; competes with Wordcab for the same regulated contact-center AI budget.

Broad incumbents

• AWS Transcribe / Call Analytics: Hyperscaler ASR service with optional Call Analytics, used as the primary cost benchmark in Wordcab's own ROI comparison; dominant incumbent with broad enterprise cloud contracts.
• Google Cloud Speech-to-Text: Hyperscaler ASR service deeply integrated with Google Cloud, often bundled with CCAI and Contact Center AI; competes with Wordcab for enterprise speech workloads.
• Microsoft Azure AI Speech: Hyperscaler ASR/TTS service inside the Azure AI stack with healthcare- and enterprise-grade certifications; competes with Wordcab for regulated enterprise and contact-center buyers.

Emerging players

• Picovoice: On-device and private-cloud voice AI for privacy-sensitive workloads; competes on the privacy-by-design axis that Wordcab also targets, though primarily at edge/on-device rather than enterprise server.
• Hume AI: Emotionally intelligent voice AI platform with custom voice deployment; emerging competitor in the private voice AI category with overlapping capabilities on TTS, voice cloning and conversational agents.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat6 records

Key risks6 records

Key highlights6 records

Customer concentration

• LinkedInlinkedin.com/company/wordcab

Compliance3 records

Revenue estimate

Valuation estimate

Number of profiles

Profiles2 records

Funding overview

Funding rounds4 records

Investors2 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Wordcab