Canadian voice-AI company providing on-device speech recognition and voice-interface technology.
Operating context
Picovoice supplies on-device components for wake-word detection, speech activity, transcription, intent recognition and speech synthesis. Porcupine and Rhino can serve Cortex-M-class embedded applications, while complete transcription and generative pipelines have different host requirements. Arm's solution brief describes speech-to-intent without an intermediate transcript. Picovoice also offers picoVLM for local image interpretation, visual question answering and OCR, with Android, C, iOS, Python and Web SDKs. This separates small embedded voice interfaces from more demanding local multimodal applications.
Canadian footprint
Picovoice Inc. is incorporated in British Columbia, and Arm identifies the business as Vancouver-based. Its Canadian contribution is on-device speech technology and embedded voice software.
Montreal, Quebec is the published city-level context for Picovoice. It does not imply a street address or exact facility location.
Picovoice provides local wake-word, speech-activity, speech-recognition, speech-to-intent and speech-synthesis components. Selected Porcupine and Rhino workflows support Cortex-M microcontrollers, while fuller voice-agent pipelines require a host suited to their individual model and runtime needs.
The publicly described On-device voice AI offering may be relevant to monitoring or remote-operation use cases; this is a derived mapping for reviewer assessment.
Reviewed connections indicate possible relevance based on public evidence. They do not indicate procurement direction, eligibility, endorsement or customer interest.
Questions for a first conversation
01
What compute, memory and licensing requirements apply to the exact offline voice or vision pipeline?
Cortex-M support for selected voice components does not imply that the full speech and vision portfolio runs on the same device.
Original sources 5 linked records
These records support the organization, its owned capabilities and their reviewed connections. A source count is not a count of independent confirmations.
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: Profile · operating context
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · technology readiness level
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · technicaldomainslugs
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · technical tags
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · core features
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · maturity
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · core features
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · commercial availability
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · technicaldomainslugs
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · defence applications
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · technical tags
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · core features
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · capability type
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · summary
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · technical tags
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · name
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · core features
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · technicaldomainslugs
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · technical tags
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Supports: On-device voice AI · defence applications
Picovoice documents local wake-word, voice-activity, transcription, intent and speech-synthesis components that support complete local pipelines or hybrid architectures. Porcupine wake-word and Rhino intent can target Arm Cortex-M; full transcription, reasoning and speech synthesis have different host requirements.
Picovoice's public evidence establishes its current role and offering, but it does not answer every access, configuration, eligibility, maturity or buyer question needed for a first business-development conversation.
Arm's partner solution brief identifies Picovoice as Vancouver-based and describes offline voice recognition on Cortex-M microcontrollers. Its speech-to-intent approach maps audio to a domain-specific command without an intermediate transcription, with models created through Picovoice Console and deployed using Shepherd.
Supports: Profile · operating context
Arm's partner solution brief identifies Picovoice as Vancouver-based and describes offline voice recognition on Cortex-M microcontrollers. Its speech-to-intent approach maps audio to a domain-specific command without an intermediate transcription, with models created through Picovoice Console and deployed using Shepherd.
Picovoice's terms identify Picovoice Inc. as a British Columbia corporation. The March 2026 terms provide a stronger legal-identity anchor than the existing unsupported Montréal location.
Supports: Profile · canadian footprint
Picovoice's terms identify Picovoice Inc. as a British Columbia corporation. The March 2026 terms provide a stronger legal-identity anchor than the existing unsupported Montréal location.
Picovoice now also offers picoVLM for local image interpretation, visual question answering and OCR. The product page lists Android, C, iOS, Python and Web SDKs and CPU execution, extending the portfolio beyond speech interfaces.
Public sources cited · Facts and assessments kept separate · Human review
The official capability material supports the named On-device voice AI offering and the bounded operational description prepared for reviewer consideration.
True North Map derives this first-conversation question from the documented Picovoice capability, current evidence and the decision gap that remains.
Supports: Profile · description
The official company material describes Picovoice as a Canadian technology business with the public offering summarized in this candidate.
Supports: Profile · canadian footprint
The official capability material supports the named On-device voice AI offering and the bounded operational description prepared for reviewer consideration.
Supports: On-device voice AI · summary
The official capability material supports the named On-device voice AI offering and the bounded operational description prepared for reviewer consideration.
Supports: On-device voice AI · summary
Arm's partner solution brief identifies Picovoice as Vancouver-based and describes offline voice recognition on Cortex-M microcontrollers. Its speech-to-intent approach maps audio to a domain-specific command without an intermediate transcription, with models created through Picovoice Console and deployed using Shepherd.
Continue exploring
Follow the strongest connections
Continue through related organizations, reviewed mission and Defence need connections, programme pathways, and explicitly linked intelligence. Similarity results describe shared areas of work, not partnerships or endorsements.
Organizations
Profiles that provide the strongest reviewed or discovery-led continuation.
