Frequently asked questions

LPS is on-device meeting intelligence. You record a conversation, and LPS produces a structured report - transcript, speaker separation, a summary, key points, and optional voice-based emotional cues - running the whole pipeline locally on your hardware.

No, and no. Transcription, speaker separation, and analysis all happen offline, on the device itself. Your audio and the resulting report never leave your hardware unless you explicitly choose to export them. You can run LPS fully air-gapped.

One pipeline runs end to end: import or record the audio, transcribe it, separate who spoke when, then analyze the transcript to produce a summary, action items, and highlights. The output is a clean, shareable report.

A typical report includes:

• A speaker-labeled transcript of the conversation.
• A concise summary and the key points discussed.
• Action items and decisions, where present.
• Optional voice-based emotional cues, presented as flagged hypotheses rather than verdicts.

LPS supports transcription and analysis in multiple languages, and can translate a report into another language for sharing. Language packs are managed on-device, so you install only what you need. Tell us the languages your team works in and we will confirm coverage for your pilot.

Entirely on your device. There is no LPS server in the processing path, so there is nothing for us - or anyone else - to intercept. Reports are stored locally and encrypted at rest.

That is exactly what it is built for. Because audio and transcripts never touch a network, LPS fits healthcare, legal, research, and other high-trust settings where sending recordings to a cloud service is not an option. It supports the traceability that supervised work requires.

No. We cannot - the pipeline runs on your side and does not phone home with your content. Any account or usage information is limited to metadata you choose to sync, never the audio or the report.

Either the device's own microphone, or a dedicated voice recorder that you plug in to import the audio. Both feed the same on-device pipeline.

We support a range of USB voice recorders and maintain a short list of models we have tested and recommend. See the supported recorders page, or order one through us.

Recordings are lightweight, and processing runs at roughly real time on a modern device. The optional advanced analysis models are a larger, one-time download that you enable only if you need them, over Wi-Fi.

As a mobile app on a dedicated device for most teams, with on-premise options for organizations that need them. In every model, processing stays local. Talk to us about the fit for your environment.

Yes. We run short pilots so you can prove the value on your own recordings before committing. Get in touch to set one up.

The app is offered in tiers that scale with what you use - from transcription only up to full analysis - billed monthly or yearly. Because processing is on-device, there are no per-token usage fees. See pricing for the current plans.