HiveRecall: we built it so words don't stay just words
As a remote team, a lot of our decisions, blockers, ideas and next steps happen during calls. We wanted that context to survive the meeting, and the tools we tried never quite gave us the mix of functionality, flexibility and privacy that internal conversations need. So we built our own: designed, built and tested in house, and today it is part of our everyday workflow.
CallNotetaker sits beside the call: pick screen or audio-only capture, choose the window and microphone, and start recording without leaving the meeting.
The library lists every recorded call with status, duration and platform, so a meeting is findable after the window closes.
Each meeting opens to audio playback, speaker timelines and a diarised transcript on Transcript, Summary and Chat tabs.
Speakers ties voice samples to names, emails and tags so recurring participants are recognised across calls.
Tags classify calls by purpose (discovery, sprint planning, support escalation) so the library filters by how the meeting was used.
The dictionary holds company names, acronyms and corrections so transcription stays accurate on internal vocabulary.
Chat settings pick the LLM provider and model, including local Ollama, so summarisation and follow-up stay under our control.
- Internal productivity for a remote software team
- Internal product, designed, built and tested in house, now in daily use
- CallNotetaker for recording, speaker recognition and transcripts, plus internal automations and MCP-based workflows that turn conversation into work
As a remote team, a lot of our decisions, blockers, ideas and next steps happen during calls, and that context used to disappear when the meeting ended. The tools we tried never quite gave us everything we needed, especially the mix of functionality, flexibility and privacy that internal conversations require. So we designed, built and tested it ourselves.
CallNotetaker records calls, recognises speakers, creates transcripts, and helps turn conversations into actual work. Internal automations and MCP-based workflows then take over, so important points become action items, blockers or backlog tasks instead of getting lost in meeting notes.
It also says how we work. We care as much about clear communication inside the company as we do with our clients, because fewer missed details and faster decisions make the client work smoother too. Building products is not limited to the products we are hired to build. If we see something in our process that can work better, we try to improve it.
HiveRecall is in daily use on our own calls. Words from a meeting become work rather than notes nobody opens again.
The deeper claim is the one about the team. We build for ourselves with the same seriousness we bring to client work, because a remote team that keeps its own context is a team that misses less for the client too. No performance metrics are published and none appear here.