Urgent.News

What's breaking now, across thousands of outlets.

Tech

How to Search Across Every Meeting Transcript You Own (Not Just the Last One)

"Someone mentioned pricing in a call last month. Which call? Who said it?" If your transcripts live as loose files from three different tools, that question takes twenty minutes. Here's how to make it take twenty seconds. Why this is hard by default Each tool searches only inside itself. Fathom searches Fathom, Otter searches Otter. Since September 2026, a free Fathom account's cross-meeting…

Discovering the right words from your past meeting transcripts can be a time-consuming task, especially when they are scattered across various platforms. To streamline this process, it is crucial to gather all your meeting transcripts into a single folder and convert them into plain text format. By doing so, you enable your computer's search function to quickly locate specific information within the files.

While folder searches may suffice for small collections of meetings, they become increasingly unreliable as the number of transcripts grows. At this stage, you'll need a system that indexes your transcripts with relevant metadata such as titles, dates, sources, speakers, and tags. This allows for more targeted searches, enabling you to filter results based on specific criteria like a particular speaker or a date range.

One effective solution is to use a tool like MeetScribe, which I developed. This software enables you to import transcripts from various formats and sources, such as TXT, VTT, SRT, and JSON, into a single, organized library. The tool then creates a local full-text index, which facilitates efficient searching and filtering without the need for an account or online uploads.

Additionally, MeetScribe offers a local web view for easier browsing of your transcripts and includes a feature to generate commitment reports across all meetings, a capability that folder search simply cannot provide.

The process is straightforward: import your transcripts once using the command `python -m meetscribe import ~/all-my-transcripts`, and then utilize the search function with filters as needed, such as `python -m meetscribe search pricing --speaker Priya --date-from 2026-10-01`. This approach saves you considerable time and effort, especially as your collection of meeting transcripts expands.

You can start with a free trial by importing your first five meetings, after which a one-time fee of $29 grants you full access. You can learn more and sign up at https://solopreneur87.gumroad.com/l/mrysy.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in Tech

The agent loop is 32 lines. Stopping it is the product.

A working agent loop is almost embarrassingly short. In Python it was about 32 lines: call the model, if it asks for a tool you run the tool, append the result, loop.

  • The agent loop comprises approximately 32 lines of code.
  • Issues like temperature settings, tool errors, and budget limits can disrupt the loop.
  • Safeguards such as human intermediaries and exit point checks are crucial for reliability.

How to Get Action Items With Owners and Due Dates Out of Your Meeting Transcripts

Every meeting ends with commitments scattered through conversation: "I'll send it by Friday." "Priya will draft the page by the 10th." "We decided to launch in January." A week later, nobody can find…

  • Manual method identifies key phrases like "I'll", "by Friday", "action item" in meeting transcripts.
  • Both methods generate detailed action items with owners and due dates from meeting transcripts.

CSS Pseudo-Classes & Pseudo-Elements: Features That Make Websites Feel Interactive

Have you ever hovered over a button on a website and watched it change colour? Maybe a navigation link suddenly became highlighted when your cursor moved over it.

  • Pseudo-classes define state-based styles (e.g., :hover, :focus)
  • Pseudo-elements target specific parts (e.g., ::before, ::after)
  • They enable interactive effects without JavaScript

NyxJ 2.0.0

NyxJ 2.0.0 marque un tournant : le projet passe d'un simple parseur à une plateforme complète pour le traitement JSON et les services web, écrite en C11 sans dépendance runtime.

  • NyxJ 2.0.0 marks significant evolution from JSON parser to comprehensive platform
  • Thread-safe parsing engine adheres to RFC 8259 with security protections
  • Includes 12 modules, web framework, CLI toolset, and extensive documentation

Amutable Foundations: A Technical Overview

Lennart Poettering, Michael Vogt, and Joaquim Rocha have put together a detailed overview of what the Amutable Operating System (AOS) and associated infrastructure will look like.

  • AOS lacks UNIX shell and SSH access, reducing attack surface
  • AOS defaults to not installing Python and Perl interpreters
  • AOS only activates signed images with Verity data

More from Saturday 7 November →