For researchers and analysts

Every interview private,
every recording searchable.

No participant data breaching IRB. No upload to a third-party SaaS your university hasn't vetted. No "they had a breach" conversation with your ethics board six months from now.

Kajo Voice research library with folders of transcribed interviews — Build a local archive you can search and export without uploading.

Oral historians, ethnographers, and field linguists run the same loop: larger archives, more languages, and deed-of-gift consent that bars cloud tools.

The jobs

File the study, then ask for the receipt

Import consented interviews into a study folder named by participant ID, transcribe on-device, pin the folder, and ask for the exact words on a topic you cannot name with a keyword. You get two to five verified quotes with speaker, timestamp, and jump-to-time — or an explicit *not in the archive*.

Synthesize the corpus

Ask what this set of participants said about a theme. The Themes starter answers with how many recordings mentioned it — a true count, never a percentage — where they agree, disagree, and go quiet, with a quote for every claim. If some interviews are still being indexed, it says so.

Brief the next interview

Before a repeat session, the Brief starter turns that participant's folder into a one-page briefing: open threads, promises, contradictions, and questions the tape still does not answer. Every bullet cited.

Attribute speakers, then code from evidence

Rename participant and interviewer after automatic speaker detection. Ask questions scoped to one speaker's turns. Export QDPX (REFI-QDA) for NVivo, ATLAS.ti, or MAXQDA when the study still needs a CAQDAS — chat is the complement, not the codebook.

Draft a findings paragraph from citations

The Draft starter writes a first-draft findings paragraph that stays tied to the tape, with inline citations and a visible review-required line. Export the answer with its quote list to Markdown or DOCX.

Erase a participant

Withdrawn consent or retention expiry: deleting the recording removes the app's audio copy, the transcript, the search index, and every chat that only used it, and strips its citations from the rest. The original file on your disk is yours to remove.

What you can actually do with it

Semi-structured interviews

Import in-person and remote interview recordings. Transcripts stay on your laptop — ready for coding without a cloud upload.

Qualitative coding

Export transcripts to NVivo, Atlas.ti, or MAXQDA. Or stay in Kajo: file by folder, search the corpus, and ask cited chat for themes.

Fieldwork notes

Import field recordings from your recorder. Audio is transcribed on your laptop — no cloud round-trip to Otter while you are in a participant’s home.

UX usability sessions

Import think-aloud session recordings, file them by study folder, and ask cited chat to surface friction moments across the corpus.

Expert interviews for market research

Import a 60-minute expert call, transcribe on-device, file by competitor or segment, and build the report from quotes you can cite verbatim.

A research workflow that stays local

Open Kajo and drop in your interview recording; name it with the participant ID. Audio is transcribed locally.
Transcription runs on your laptop. No upload.
Put the recording in a study folder. Folders are the org axis — drill in, search within a project, keep fieldwork tidy without a cloud repository.
Open Chat in Kajo and ask “What did participants say about trust?” — pin the study folder as scope; every answer cites verbatim quotes from your transcripts.
Export transcripts as plain text, JSON, CSV, or QDPX for NVivo / Atlas.ti / MAXQDA — or keep working in Kajo with search and cited chat.

IRB / Ethics

Kajo’s architecture aligns with how research ethics boards already think about data-handling: audio is captured and processed on the researcher’s hardware, transcripts are stored at rest on that same hardware, and no third-party SaaS is involved — audio never leaves the device.

Storage and retention. Every capture is a file on your disk. Your IRB protocol’s retention policy is enforced by you, not by a vendor’s deletion API. Delete a participant’s recording from the Library — that removes the transcript, the search index, and the chats that only used that recording. The original audio file on disk stays where it is; remove it yourself when your protocol says so.

Right to erasure. Participants who withdraw consent: you delete their file. No cloud delay, no “we will purge from backups within 30 days” footnote. The data is on your laptop; it goes when you put it in the trash — the original file on disk is yours to remove.

Export and portability. Per-file export (TXT, JSON, CSV, DOCX, MD, PDF, QDPX, and more) is available on Free within the allowance. Lifetime unlocks batch export when you need to ship many transcripts at once. There is no separate cloud export API — your corpus stays on your laptop.

Sample IRB language. Email hello@kajovoice.com and we will send a paragraph you can drop into your protocol describing Kajo’s data flow. We are not lawyers; your IRB chair is.

Tool stack

Stage	Tool	What it does
Capture	Kajo desktop	macOS, Windows, Linux. On-device ASR.
Query	Kajo cited chat	Built-in local model. Ask cross-corpus questions about your participant data with source citations.
Coding suite	NVivo / Atlas.ti / MAXQDA	Export QDPX, TXT, or DOCX from Kajo; code and cite in your CAQDAS tool.

Three reasons it fits research workflows

Local-first / private by default

Your audio never leaves your laptop.

Audio is transcribed on-device — there is no cloud-processing path. Private by architecture.
Cross-platform, including Linux

Mac, Windows, and Linux — one product, three operating systems.

Native installers for all three. Linux is a first-class target, not a port — the only local audio knowledge base that takes it seriously.
Cited chat

Ask questions across your archive — get cited answers.

Chat with your entire library using a local model. Every answer links back to the exact moment in the recording — no copy/paste.

Free

5 files or 150 minutes of audio — total, for life. Core loop (transcribe, translate, summarize, chat), no card, no upload.

Transcribes 98 languages on your laptop — all on-device.
On-device translation and summaries — no cloud-processing path
Private searchable library — full-text + semantic search
Folder organization
Cited chat across your archive
Per-file export (SRT/VTT/TXT/MD/DOCX/PDF/QDPX and more)
Mac, Windows, and Linux

Lifetime — $49 once

Unlimited files and minutes
Unlimited retrieval index
Watch folders — auto-import new recordings
Batch export
Up to 3 machines at once

Your installed version keeps working forever.

Frequently asked questions

Does Kajo satisfy my IRB's data-handling requirements?

Kajo runs transcription on-device — no audio leaves the laptop. Your IRB protocol will need to disclose that participant audio is processed locally and stored at rest on your hardware. Retention and right-to-erasure: delete a participant's recording from the Library — that removes the transcript, search index, and chats that only used that recording. Chats that search the entire library stay, with citations to the deleted file stripped. The original audio file on disk is not deleted. We have a sample IRB language template — email hello@kajovoice.com.

Can I export transcripts for coding in NVivo / Atlas.ti / MAXQDA?

Yes. Per-file export on Free includes QDPX (REFI-QDA) plus TXT, JSON, CSV, DOCX, MD, PDF, and caption formats — ready for NVivo, Atlas.ti, and MAXQDA. Lifetime unlocks batch export when you need to ship many files at once. You can also export with sentence-level timestamps for line-by-line coding.

How is this different from Otter.ai or Trint for academic research?

Otter and Trint upload your participant audio to the cloud. For interviews with vulnerable populations, that is a non-starter under most IRB protocols and university privacy policies. Kajo transcribes on-device — no third party touches the audio. If your IRB has approved Otter or Trint specifically, your IRB approval for Kajo will typically be easier because the data-flow story is shorter.

I do mixed-methods. Can I organize interviews and observations together?

Yes. Use nested folders as the org axis — e.g. a study folder with interview and observation subfolders — then pin that folder as chat scope to ask across the whole corpus.

All roles QDPX export without the cloud vs Dovetail vs Skimle Trust Center

Your fieldwork. Your laptop. Your data.

Download Kajo, run a pilot interview, and watch the transcript appear without touching the cloud.

Every interview private, every recording searchable.