Hyper – A stupidly non-corporate voice AI app for IRL conversations
Hey HN, we're building Hyper: an iOS voice recorder for the conversations that matter too much to pause and take notes. 1:1s with your cofounder, coffee chats, standups, whiteboard sessions. One tap to record, and when you're done you get a summary, full transcript, and tappable action items (calendar events, reminders, email drafts) — no notes needed. We built this because at our last company (a robotics startup), the most consequential decisions happened spontaneously — in hallways, over lunch, on walks. None of it got captured. Every recorder we tried was built for scheduled Zoom calls…
In plain words
Hyper is an iOS voice recorder designed for capturing unstructured conversations like one-on-one meetings, coffee chats, and brainstorming sessions. It records conversations with a single tap and automatically generates summaries, full transcripts, and actionable items such as calendar events, reminders, and email drafts. Users can ask questions during recording using voice commands without interrupting the conversation. The app targets professionals and teams who need to preserve spontaneous, consequential discussions that happen outside of scheduled video calls.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
Hey HN, we're building Hyper: an iOS voice recorder for the conversations that matter too much to pause and take notes. 1:1s with your cofounder, coffee chats, standups, whiteboard sessions. One tap to record, and when you're done you get a summary, full transcript, and tappable action items (calendar events, reminders, email drafts) — no notes needed. We built this because at our last company (a robotics startup), the most consequential decisions happened spontaneously — in hallways, over lunch, on walks. None of it got captured. Every recorder we tried was built for scheduled Zoom calls with agendas. We wanted something for the messy, unstructured conversations where the real work happens. How it works: Hyper transcribes your conversations in real-time. On-device wakeword detection lets you ask it questions mid-conversation with “hey hyper” or via chat without breaking flow — "wait, what did we decide about the deadline last time?" — and it pulls from every past conversation. After the meeting, another agent extracts structured actions and maps them to the rest of your tools (like Calendar, GMail, and native iOS integrations). The bet we're making: existing meeting tools stop at "here's a document." We believe people don't want better documents from their meetings, they want answers to their questions. We envision a future where people can simply ask across all conversations and get the answer instantly — without files or folders or manual organization. We’re building out a long-term memory framework that lets Hyper resolve conflicting information, surface stale information, and help untangle the gnarliest of problems. We're two founders, currently self-funded, building in SF. We’d love for you to try it out in your next exciting conversation and to hear your feedback. Thanks! -- Shalin & Kanyes
Does the same job
all alternatives →
- IBI built an app for when I talk too much in online meetings2022 · unblah.me · ▲451
Hey HN! Alexis here, I’m a product manager and software developer in Berlin by way of New York. I want to show you this app I made – It’s like a "buddy" for those, like myself, who inadvertedly talk too much in meetings. The app gives me feedback and a little more in control of what I have influence over by: * Keeping track of how long I’ve been speaking * Catching myself before I talk too much * Developing a better sense of timing I truly love having conversations with people in real-life. But online meetings, especially group calls, tend to make me nervous. I can't read body language. The…
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 27d ago · cactuscompute.com
Launched alongside, March 2026
the whole month →
Switch from ChatGPT to Claude with import memory feature
AI · Mar 2026 · claude.com
