PRESS RELEASE 2026.02.17

TIGEREYE Launches Free Beta of "SecureEdge LLM & Voice Assist," a Fully Offline AI App that Completes Recording, Transcription, and Summarization on Smartphones


On February 17, 2026, in Tokyo — TIGEREYE Inc. (Headquarters: Chuo-ku, Tokyo; CEO: Gaku Kamimura) announces the launch of a free beta version of "SecureEdge LLM & Voice Assist," an AI app that completes recording, transcription, and summarization entirely on smartphones using local LLM and local speech recognition.

Designed without reliance on external server communication, the app keeps sensitive information on-device. It is suitable for cloud-restricted environments and confidential meetings where privacy and compliance are critical.

SecureEdge LLM & Voice Assist free beta image

Local LLM (Qwen) × Local Speech Recognition × Fully Offline

SecureEdge LLM & Voice Assist transcribes audio locally from meetings, sales conversations, and interviews, then uses the local LLM "Qwen" to generate concise summaries.

By completing all processing on-device, it enables stable operation without network dependency while maintaining strict data confidentiality.

Telemetry-Free, Fully Local Operation

The app does not collect usage telemetry. From launch to exit, audio, transcription, and summary data are not transmitted externally.

This design provides the following benefits:

  • • Fully offline operation: Works without an internet connection
  • • Privacy protection: Recording, transcription, and summary data remain on-device
  • • Fit for restricted networks: Suitable for enterprises and medical institutions with strict policies
  • • Zero communication cost: No mobile data usage required

What is Qwen?

Qwen is an open-source large language model (LLM) developed and released by Alibaba Cloud. It supports multiple languages and offers strong Japanese-language processing. In this app, Qwen runs directly on iPhone to provide high-accuracy summarization without cloud connectivity.

Key Features

Recording One-tap recording, waveform display, and auto-record settings
Transcription On-device speech-to-text conversion via local speech recognition
Summarization Bullet-point summaries using local LLM "Qwen," with custom prompt support
Speaker Identification Distinguishes multiple speakers such as meeting participants
Security Fully offline processing, no external communication, no telemetry

Expected Use Cases

[General Business]

  • Meetings: Automated minutes and summary of key points
  • Sales: Record customer conversations and streamline reporting
  • Interviews: Accurate capture and transcription of statements
  • Training/Seminars: Record sessions and organize highlights
  • Confidential meetings: Safe recording without external communication

[Medical and Care Settings]

  • Home care visits: Record reliably regardless of network conditions
  • Clinical note support: Reduce post-consultation documentation time
  • Informed consent records: Preserve explanations accurately
  • Team coordination: Share summaries for efficient handovers

[Restricted Network Environments]

  • Confidential departments: Use AI even where external communication is restricted
  • Offline sites: Keep records where no communication infrastructure is available
  • Overseas business trips: Use anywhere without roaming concerns
    * Downloading the initial model requires network communication.

Offering Overview

App Name SecureEdge LLM & Voice Assist
Supported Device iPhone (iOS 17 or later)
Availability Beta version, free of charge
Launch Date February 17, 2026

Future Outlook

TIGEREYE will continue improving the product based on feedback from users across industries. Planned enhancements include external system integration, multilingual support, and industry-specific templates.

This app introduces a practical approach to AI adoption: improving operational efficiency while upholding the essential requirement that sensitive information remains on-device. By eliminating all external communication, including telemetry, it delivers a true privacy-first architecture.


Back to List

Please feel free to contact us.

Contact us