<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Kajo Voice Blog</title><description>Transcription, AI, and productivity insights from the developer of Kajo Voice</description><link>https://kajovoice.com/</link><language>en-us</language><item><title>Arabic Transcription Software: MSA, Dialects, and Code-Switching</title><link>https://kajovoice.com/blog/arabic-transcription-software/</link><guid isPermaLink="true">https://kajovoice.com/blog/arabic-transcription-software/</guid><description>How Kajo Voice handles Modern Standard Arabic, Egyptian, Gulf, Levantine, and Moroccan dialects — plus Arabic-English and Arabic-French code-switching — with on-device transcription, speaker labels, and translation.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>What Reads This Number Next? A BM25 Defect That Dissolved</title><link>https://kajovoice.com/blog/bm25-length-normalization-rrf-rank/</link><guid isPermaLink="true">https://kajovoice.com/blog/bm25-length-normalization-rrf-rank/</guid><description>I filed a defect against my own search engine straight from the BM25 formula. Measured, the knob I proposed did not exist, the penalty was 25% rather than 40%, and within one script the length penalty cannot move the ranking at all — because fusion reads rank and throws the scores away.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>My Chunker Counted Characters and Called Them Tokens</title><link>https://kajovoice.com/blog/chunk-sizing-tokenizer-not-characters/</link><guid isPermaLink="true">https://kajovoice.com/blog/chunk-sizing-tokenizer-not-characters/</guid><description>The function deciding where every chunk of every transcript ended was a characters-per-token estimate nobody had compared with a tokenizer. It read 0.96 of the truth in Latin script and 0.48 in Devanagari — and the multilingual folder scoring 12 of 12 held ten documents of one chunk each.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Code-Switching in Transcription: How AI Handles Multilingual Speech</title><link>https://kajovoice.com/blog/code-switching-transcription-guide/</link><guid isPermaLink="true">https://kajovoice.com/blog/code-switching-transcription-guide/</guid><description>What code-switching is, why it breaks most transcription tools, and how Kajo Voice&apos;s multilingual on-device transcription handles mid-sentence language transitions — honestly, including what to review.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Free Transcription Software in 2026: What &quot;Free&quot; Gets You When the Audio Can&apos;t Be Uploaded</title><link>https://kajovoice.com/blog/free-transcription-software/</link><guid isPermaLink="true">https://kajovoice.com/blog/free-transcription-software/</guid><description>A roundup of transcription tools that are free and keep audio on your machine — manual tools, the open-weight model from a terminal, local desktop apps, and Kajo Voice Free — with the cloud free tiers as the counter-example.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>HIPAA and Local Transcription: Keeping PHI on Your Device</title><link>https://kajovoice.com/blog/hipaa-and-local-transcription/</link><guid isPermaLink="true">https://kajovoice.com/blog/hipaa-and-local-transcription/</guid><description>How on-device transcription changes the HIPAA conversation for therapists, clinical supervisors, and other session-based practitioners who already record: when protected health information never leaves your machine, there is no transcription vendor in the chain.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>How to Transcribe an Interview Archive on Your Own Machine</title><link>https://kajovoice.com/blog/how-to-transcribe-audio/</link><guid isPermaLink="true">https://kajovoice.com/blog/how-to-transcribe-audio/</guid><description>A step-by-step guide to turning the recordings you already have into transcripts, speaker labels, summaries, and a private library you can ask with citations — on your own machine, with no upload. Formats, review, export, and what to expect.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Kajo Voice vs the Whisper CLI: When to Use Each</title><link>https://kajovoice.com/blog/kajo-vs-whisper-cli/</link><guid isPermaLink="true">https://kajovoice.com/blog/kajo-vs-whisper-cli/</guid><description>Whisper from a terminal is powerful and free. Kajo Voice runs the same open-weight model inside a desktop app with speaker labels, a private library, on-device translation and summaries, and cited chat. Here&apos;s how they compare, honestly.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Linux Desktop Transcription: Local AI Without a Mac Tax</title><link>https://kajovoice.com/blog/linux-desktop-transcription/</link><guid isPermaLink="true">https://kajovoice.com/blog/linux-desktop-transcription/</guid><description>Most voice AI apps skip Linux. Kajo Voice ships the same local-first app on Linux as on Mac and Windows — on-device transcription, speaker labels, a private library, and cited chat — as .deb, .rpm, and AppImage for x86_64 and ARM64.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Cited Chat Over Local Audio Archives: Ask Your Recordings, Keep the Receipts</title><link>https://kajovoice.com/blog/local-cited-chat-over-audio-archives/</link><guid isPermaLink="true">https://kajovoice.com/blog/local-cited-chat-over-audio-archives/</guid><description>How on-device cited chat turns a private transcript library into a searchable knowledge base — answers linked back to the source recording, scoped to the folder you are in, with no cloud upload.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Local vs Cloud Transcription: Which Is Right for You?</title><link>https://kajovoice.com/blog/local-vs-cloud-transcription/</link><guid isPermaLink="true">https://kajovoice.com/blog/local-vs-cloud-transcription/</guid><description>A complete breakdown of local AI transcription (Kajo Voice) vs cloud services (Otter, Fireflies, AssemblyAI). Privacy, cost, accuracy, offline use, and compliance — all compared honestly, with sourced prices.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Multilingual Transcription Software: Every Language in Your Recordings, Including Code-Switching</title><link>https://kajovoice.com/blog/multilingual-transcription/</link><guid isPermaLink="true">https://kajovoice.com/blog/multilingual-transcription/</guid><description>Kajo Voice transcribes dozens of languages on your own machine and keeps mixed-language speech as spoken — Spanish↔English, Arabic↔French — then translates on-device. No cloud, no per-minute bill.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Offline AI Transcription: Download Once, Then Disconnect</title><link>https://kajovoice.com/blog/offline-ai-transcription/</link><guid isPermaLink="true">https://kajovoice.com/blog/offline-ai-transcription/</guid><description>How Kajo Voice runs the whole loop — transcription, speaker labels, translation, summaries, cited chat — with the network off on Mac, Windows, or Linux: what downloads once, and what still talks to the network when it can.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Offline Transcription on Windows: Private, On-Device, and Searchable</title><link>https://kajovoice.com/blog/offline-transcription-windows/</link><guid isPermaLink="true">https://kajovoice.com/blog/offline-transcription-windows/</guid><description>How to transcribe audio and video privately on a Windows 10 or 11 PC — on-device, offline after setup, with speaker labels, a searchable library, and answers you can check. What it needs, how fast it is, and what never leaves the machine.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>One-Time vs Subscription Transcription: The Real Cost of Owning Your Archive</title><link>https://kajovoice.com/blog/one-time-vs-subscription-transcription/</link><guid isPermaLink="true">https://kajovoice.com/blog/one-time-vs-subscription-transcription/</guid><description>Compare subscription transcription pricing to a one-time lifetime purchase — why local-first economics favor paying once for on-device transcription, a private library, and cited chat.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>An Otter.ai Alternative That Never Uploads Your Audio</title><link>https://kajovoice.com/blog/otter-alternative-that-never-uploads/</link><guid isPermaLink="true">https://kajovoice.com/blog/otter-alternative-that-never-uploads/</guid><description>Looking for an Otter alternative without cloud uploads? Compare local-first Kajo Voice — on-device transcription, speaker labels, a private library, cited chat — to Otter&apos;s upload-and-subscribe model.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>The Optimisation I Had Already Rejected, Twice</title><link>https://kajovoice.com/blog/prompt-lookup-decoding-rag-citations/</link><guid isPermaLink="true">https://kajovoice.com/blog/prompt-lookup-decoding-rag-citations/</guid><description>Forty-two percent of my answer time went on the model retyping text out of its own prompt, so I wired up the technique built for exactly that. It came out 44% slower. The payoff rides on one ratio — about 15% in my output — and I had already measured it, days earlier, into a flag I never read.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>QDPX Export for NVivo Without Uploading Your Interviews</title><link>https://kajovoice.com/blog/qdpx-nvivo-export-without-cloud/</link><guid isPermaLink="true">https://kajovoice.com/blog/qdpx-nvivo-export-without-cloud/</guid><description>How qualitative researchers get REFI-QDA (QDPX) exports from the local-first audio knowledge base — ready for NVivo, ATLAS.ti, and MAXQDA — without sending audio to the cloud.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>When the Evaluation Is the Bug</title><link>https://kajovoice.com/blog/rag-evaluation-measurement-bugs/</link><guid isPermaLink="true">https://kajovoice.com/blog/rag-evaluation-measurement-bugs/</guid><description>I shipped a passage splitter behind 9,929 green tests, ran the evaluation with it on and with it off, and got identical numbers both times. Three readings from one harness: a change that never ran, a corpus that counted itself twice, and my fix for that, which collapsed 300 recordings onto 24.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Spanish-English Interviews: Transcribe the Mix as Spoken, Ask in Either Language</title><link>https://kajovoice.com/blog/spanish-english-transcription/</link><guid isPermaLink="true">https://kajovoice.com/blog/spanish-english-transcription/</guid><description>Kajo Voice transcribes Spanish-English interviews on your machine as they were spoken, translates on-device, and answers questions across a folder with citations — no upload, no per-minute cloud bill.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Transcription Accuracy: What to Expect from On-Device Transcription — and How to Check It on Your Own Audio</title><link>https://kajovoice.com/blog/transcription-accuracy-comparison/</link><guid isPermaLink="true">https://kajovoice.com/blog/transcription-accuracy-comparison/</guid><description>What published benchmarks say about the model Kajo Voice runs, what actually drives word error rate, how to get a better recording without a cloud cleanup step, and how to measure accuracy on your own audio before you pay.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Transcription for Journalists: Source-Protecting by Construction</title><link>https://kajovoice.com/blog/transcription-for-journalists/</link><guid isPermaLink="true">https://kajovoice.com/blog/transcription-for-journalists/</guid><description>Journalists need transcription that is searchable and protects sources. Kajo Voice runs on your machine — no audio uploaded, no third party with access to your interviews, and a private archive you can ask with citations.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Transcription for Qualitative Researchers: Interviews, Focus Groups, and Field Recordings</title><link>https://kajovoice.com/blog/transcription-for-researchers/</link><guid isPermaLink="true">https://kajovoice.com/blog/transcription-for-researchers/</guid><description>How academic and UX researchers use Kajo Voice — the local-first audio knowledge base — to transcribe interviews on their own machines, label speakers, ask a study folder with citations, and export QDPX.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Kajo Voice: قاعدة المعرفة الصوتية المحلية — نسخ وترجمة ودردشة باقتباسات على جهازك</title><link>https://kajovoice.com/blog/transcription-locale-ar/</link><guid isPermaLink="true">https://kajovoice.com/blog/transcription-locale-ar/</guid><description>Kajo Voice يحوّل أي تسجيل صوتي أو مرئي إلى نص ومكتبة خاصة قابلة للبحث على جهازك — دون رفع أي ملف إلى الإنترنت. يدعم 99 لغة بما فيها العربية، مع تمييز المتحدثين والترجمة والدردشة باقتباسات. متاح لـ Mac وWindows وLinux.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>The Privacy Guide to AI Transcription: What Happens to Your Audio</title><link>https://kajovoice.com/blog/transcription-privacy-guide/</link><guid isPermaLink="true">https://kajovoice.com/blog/transcription-privacy-guide/</guid><description>A detailed guide to what cloud transcription services do with your audio, the GDPR and confidentiality implications, and how to choose a transcription tool that protects sensitive conversations.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Watch Folders for Interview Archives: Auto-Import Without Cloud Sync Theater</title><link>https://kajovoice.com/blog/watch-folders-for-interview-archives/</link><guid isPermaLink="true">https://kajovoice.com/blog/watch-folders-for-interview-archives/</guid><description>How researchers and journalists use local watch folders to grow a private interview archive — automatic intake on disk, on-device transcription, no upload pipeline. A Lifetime feature, explained honestly.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Whisper Large-v3 Turbo: The Model Kajo Voice Runs and Why</title><link>https://kajovoice.com/blog/whisper-large-v3-turbo-guide/</link><guid isPermaLink="true">https://kajovoice.com/blog/whisper-large-v3-turbo-guide/</guid><description>Whisper large-v3 turbo is OpenAI&apos;s speed-optimized Whisper variant — and Kajo Voice&apos;s single on-device transcription engine. Here&apos;s how it works, how it compares to the full large-v3, and why it&apos;s the right choice for on-device use.</description><pubDate>Tue, 27 Oct 2026 00:00:00 GMT</pubDate></item></channel></rss>