Meeting Notes & Transcription

Whisper is an automatic speech recognition (ASR) system developed by OpenAI. It is a state-of-the-art model trained on a large dataset of diverse audio to perform tasks such as multilingual speech recognition, speech translation, and language identification. Whisper is designed to be robust to accents, background noise, and technical language. It is available as an open-source model, allowing developers to integrate it into applications for transcription, translation, and other speech-related tasks.

Kaltura is a video platform that universities use to store, caption, stream and share lecture recordings and other course video, often inside the LMS. The vendor says its REACH service produces machine (AI) captions, translations and dubbing, and that its Genie conversational agents (Class Genie in the LMS) answer student and instructor questions from the institution's own captioned videos and documents, with video clips, flashcards and quizzes. The vendor says machine captions average 85% accuracy, that AI-generated captions and summaries carry a notice and may need human review, and that it does not use customer data to train its models. Kaltura's AI terms bar customers from using its AI for EU AI Act high-risk purposes, including educational admission or evaluation decisions.

Attendi is speech-reporting software for Dutch long-term care, mainly elderly care and disability care and also mental health care, used by nurses, carers and other care professionals. Spraakgestuurd Rapporteren is built into the electronic client record. The professional dictates on phone or web and Attendi's own speech model, which the vendor says it trained on millions of spoken care reports, writes the text with punctuation; its Schrijfhulp feature takes spoken instructions such as 'make this a SOEP report' and rewrites the text with AI. The Attendi App records longer conversations such as intakes and multidisciplinary meetings and drafts a structured summary per client from templates the organisation sets; it syncs to Nedap Ons. The vendor says the professional checks and edits the text before it is saved. Its privacy statement says care data is hosted with Microsoft Azure, audio is kept at most 24 hours and transcripts and summaries at most 30 days.

Dragon Medical One is Microsoft's cloud speech-recognition software for clinical dictation by clinicians, originally developed by Nuance and still administered through Nuance-branded tools. Clinicians speak into a microphone or a smartphone running PowerMic Mobile and the text appears in the electronic health record or a separate Dictation Box; voice commands navigate the record, format text and insert stored text blocks (auto-texts). The vendor says it offers 'best-in-class' speech recognition with automatic accent detection and audio calibration and needs no voice-profile training. The output is the clinician's own dictated words, which the clinician edits in place or in the Dictation Box before transferring it to the record; no AI summary is involved. Nuance's 2019 Dutch launch release said it is hosted on Microsoft Azure; Microsoft lists regional data centres in the US, Australia, Canada, France, Germany, Switzerland and the UK but not which serves Dutch customers. Microsoft promotes migration to Dragon Copilot but still releases Dragon Medical One (2026.3, September 2026).

Medendo is a speech-to-report assistant for care professionals, built mainly for Dutch mental health care. Psychologists, psychiatrists, nurses and other practitioners use it, from solo practices to large institutions. It runs in a browser on computer or phone and records in-person and video sessions or dictation. The vendor says AI turns the conversation into a written report in about 20 seconds, using templates for the whole treatment cycle, such as intake, diagnostics, treatment plan and session report, which users can define; it also offers an AI chat that can draw on the client record. The vendor says the clinician checks the report afterwards and nothing enters the record without their approval; reports can be written back to Medicore, Nedap Ons and SDB. The vendor says conversation data is processed and stored encrypted only in the EU, audio is deleted once transcribed, transcripts and reports are deleted after 7 days by default, and the suppliers that produce transcript and report retain nothing.

Quallie.Ai is an AI analysis platform for qualitative research. It transcribes and analyses interviews, focus groups and communities data (audio, video and transcripts), with every finding linked back to the moment it was said, keeping analysts close to the evidence. It supports transcription in 50+ languages with one-click translation to English, chat with the data, question grids and mind maps, quote-based clips/reels with auto-subtitles, and persistent research memory (living reports). Intended users include in-house enterprise research teams, agencies and independent researchers. Evidenced use cases include concept testing, journey mapping, creative & UX evaluation, new product development, ethnography & org culture, foresight & futures, CX research and comms/brand research; used by teams such as Stripe Partners, YONDER, common strategies and an insights analyst at nature.

Zoom AI Companion is the set of generative-AI features in Zoom Workplace, such as meeting summaries, chat and email compose, transcription, call summaries and questions about past meetings; in June 2026 Zoom retired the AI Companion brand, and these features now go by functional names and remain included with eligible paid Zoom Workplace plans, while ZoomMate is a separate paid agentic add-on. Zoom markets the AI features to schools and universities for lecture summaries, smart chapters, live translation and student services. The vendor says it uses a federated approach that combines Zoom-hosted models with models from Anthropic, OpenAI and Google Gemini, that it does not use customer audio, video, chat, screen sharing or attachments to train Zoom's or third-party models, and that third-party providers work under zero data retention. Zoom says data is processed and stored according to the account's hosted data-centre region, that EU-hosted customers get in-region model processing, and that admins can disable and lock features such as meeting transcripts.

Autoscriber is clinical documentation software for Dutch and Belgian care organisations, used by medical specialists, GPs, nurses, mental health and forensic care professionals, occupational physicians and allied health staff in consultations, at the bedside and in handovers. The vendor says it records the conversation, transcribes it and turns it into structured record entries such as SOEP notes, history, medication and treatment plans, which an integration can write into the electronic record; clinicians can also dictate notes. The vendor says the professional validates the output before it is filed, the AI takes no autonomous decisions and validations are logged in an audit trail. Its terms say diagnosis and treatment planning stay with the clinician and that the product is not guaranteed to be error-free. The vendor's sub-processor list names Microsoft Azure and Google Cloud for hosting and running models and Speechmatics for speech-to-text, all processing within the EEA; audio is deleted after processing.

NVivo AI Cloud is a browser-based environment for qualitative and mixed-methods research from Lumivero, sold separately from NVivo Desktop, which the vendor says has no AI features. Academic researchers, students and research teams use it to import, transcribe, code and analyse interviews, documents and literature. The vendor says its AI assistant AURA guides project set-up, summarises sources against the research questions, suggests initial topics and codes, surfaces cross-case patterns and outliers, and answers questions about the project with citations back to the sources, including a Devil's Advocate mode that looks for disconfirming evidence. Researchers confirm, edit or reject each AI suggestion before it enters the analysis, and every AI-assisted step is recorded in an exportable audit trail.

Fathom is an AI-powered meeting assistant designed to automatically record, transcribe, highlight, and summarize meetings across platforms like Zoom, Google Meet, and Microsoft Teams. It integrates with CRM systems (e.g., Salesforce) and productivity tools (e.g., Notion, Slack) to streamline workflows. Fathom also offers features like action item tracking, speaker separation, and customizable summaries. The tool is primarily targeted at professionals, sales teams, and businesses to improve meeting efficiency and documentation.