Notare maps to your existing Word template, applies your jurisdiction's formatting rules, and identifies speakers from the record. You don't learn a new system. Your system just got faster.
Designed by a working court reporter. Every feature reflects real courtroom workflow.
Generic AI transcription produces unformatted text with Speaker A and Speaker B labels. You still have to apply your template styles, fix the speaker names, build the caption page, run your proofing rules, and format the whole thing to your agency's spec.
That's not a time-saver. That's a starting point, and often not a very good one.
Notare starts where other tools stop. It outputs a formatted, styled, speaker-attributed transcript using your template, not ours.
See the transcript build in real time during the proceeding, like closed captioning. Edit words as they appear. Flag unclear sections with a hotkey and replay 20-second audio snippets for review without stopping the recording.
Every transcript goes through a multi-layer review: confidence scoring, legal terminology checks, proper noun consistency, and an AI context pass that catches mistakes the speech engine was confident about. Flags appear in a single review queue.
Built-in rules clean up informal speech, capitalize legal titles (Your Honor, THE COURT, Counsel), fix punctuation and spacing. Every rule is customizable per profile: turn them on, off, or add your own. Hover any option to see exactly what it does.
Notare uses voice embeddings to distinguish speakers by how they sound, not just by what they say. Known speakers from case documents are matched to voice clusters automatically. Multi-mic setups map channels directly to named participants. Rename one instance and every matching segment updates.
Share a live link during any session; attorneys see the transcript updating in real time from their browser. After the proceeding, generate secure share links with view, download, and print controls. Witness read-and-sign built in.
Two editors can work the same transcript at the same time. Segment locking prevents conflicts. Changes sync live. Built for a reporter and proofreader splitting the work, or two reporters on a long proceeding.
Upload your Word template once, and Notare auto-detects your paragraph styles, bookmarks, and macros. Export to court-standard deposition format (with caption, appearances, index, certificate, and errata sheet), Word (.docx/.docm with macros intact), full-size PDF, condensed PDF (4-on-1), ASCII transcript, LEF (LiveNote), PTX (Sanction), video sync (.dvt), word index, reference document, raw ASR, plain text, or JSON. All exports save locally with no browser downloads, no security flags, and no blocked macros.
Search every transcript you've ever produced from the dashboard. Find testimony by keyword, speaker, or case. Results show the matching text with context. Instant access to your complete archive.
Jurisdiction profiles store your template, formatting rules, proofing preferences, hotkeys, bookmark mappings, and certificate language. Select a profile when creating a job and everything auto-fills: court name, reporter info, styles, and rules. Set it up once per jurisdiction, use it every day.
Upload a notice of appeal, judgment, docket sheet, subpoena, or any case filing in PDF, Word, HTML, or text format. Notare identifies the document type, then uses built-in NER to extract case number, parties, attorneys, judge, dates, and times. Everything fills automatically. All processing stays on your machine.
Remote via Zoom or Teams, in-person with multi-mic hardware and channel mapping, single room mic with diarization, or pre-recorded file batch processing. Each mode includes transparent accuracy expectations.
Choose the engine that fits your workflow: Whisper (100% local, no costs), AssemblyAI, Deepgram Nova-3, or Rev.ai (built for legal). All produce the same output format. If any cloud engine fails, Notare falls back to local automatically.
Your expertise is in the room: listening, capturing, ensuring the record is complete. Notare takes the formatting, styling, and compliance work off your plate so you can focus on accuracy.
Deposition transcripts returned in hours instead of days. Case filings auto-populate caption pages. Multi-channel audio ensures every speaker is captured distinctly.
Notare runs entirely on your infrastructure. No recordings leave the courthouse. Profiles are administered centrally and enforced across all reporters in the system.
Notare's local engine processes audio entirely on your machine. Transcripts and case data are stored on your machine. When using the local Whisper engine, nothing is transmitted to external servers.
The default Whisper engine runs entirely on-premise. No audio leaves your machine. Cloud engines (AssemblyAI, Deepgram, Rev.ai) are optional and clearly labeled; you choose what stays local.
Transcripts, audio files, and case data are stored locally. You control retention, deletion, and sharing. No external accounts required for local operation.
Architecture designed with the requirements of criminal justice and sealed proceedings in mind. Local processing means sealed records stay sealed by default.
Built by court reporters, for court reporters. Try it free for 30 days.
Every tier includes unlimited local transcription. You pay for features, not usage. No one gets cut off mid-job.
The complete core toolkit
Annual: pay 10 months, get 12 ($590/yr)
Get StartedEverything in one seat, all features
Annual: pay 10 months, get 12 ($890/yr)
Get StartedUp to 5 seats, shared workflow
Annual: pay 10 months, get 12 ($1,190/yr)
Get StartedUnlimited users. Annual offline licensing. Dedicated onboarding. Phone support. Coming soon: court-administered profiles.
Contact SalesCourt reporting programs & training institutions. Batch licensing, instructor tools, student pricing.
Contact SalesTry Notare free for 30 days.
Download the installer, then request a trial key. Full Professional access. no payment, no credit card.
Automatic speech recognition (ASR) produces a first draft that requires professional review and correction before it is suitable for certification or filing. ASR errors. including mishears, speaker misidentification, and formatting artifacts. are known, recurring, and expected across all speech recognition models and vendors. This is an industry-wide reality, not specific to Notare.
Notare aims to produce a better starting point by applying your formatting rules, detecting speakers from voice characteristics, and flagging low-confidence segments for review. However, the responsibility to verify accuracy remains entirely with the professional producing the transcript. That standard does not change because the starting point improved.
Transcript quality depends directly on audio quality. Multi-channel hardware separation provides the best results. Single-microphone room capture in a noisy environment provides the worst. Speaker detection accuracy varies based on audio conditions, number of speakers, and recording setup. These factors are disclosed at job setup.
When using cloud transcription engines (AssemblyAI, Deepgram, Rev.ai), audio is transmitted to those services for processing. The local Whisper engine processes everything on your machine with no external transmission. Your choice of engine is clearly indicated in settings.
Notare is a productivity tool for professional court reporters and transcriptionists. It is not a replacement for the professionals who ensure the integrity of the legal record, and no output should be certified or filed without professional review.
Download Notare, request a free 30-day trial key, and see how it fits your workflow. No commitment, no credit card.