ATRIUMsearch → argument graph
ClaimVideo · 2:24 — 3:54

To make Reachy a true personal assistant requires wake word detection, back-and-forth conversation, and memory, tool calling, coding tasks, Home Assistant control, and calendar reading.

Scott lays out the full checklist of capabilities Reachy Mini needs to count as a genuine personal assistant, from wake-word detection to memory and calendar access. ✦ AI generated

Scott · Syntax · 2026-07-17 · original ↗

starts at this moment · 2:24

I need to, one, connect it to my Hermes. I need to get wake word detection, so that way I can talk to it directly starting out of nowhere. I need to have conversation going back and forth. And as far as the personal assistant tasks go, I would love for this thing to have a memory.

verbatim transcript · starts at 2:24

Transcript · around this moment

2:24true personal assistant and claim success here, I need this robot to do a few things. I need to, one, connect it to my Hermes. I need to get wake word detection, so that way I can talk to it directly starting out of nowhere. I need to have conversation going back and forth. And as far as the personal assistant tasks go, I would love for this thing to have a memory. I would

2:49like for it to be able to do tool calling. I would like for it to do simple coding tasks, be able to use my Home Assistant. And I would love for this to be able to do personal assistant tasks like simply reading my calendar. A lot of those are very possible with Hermes, but many of these things are going to require true custom software. Now, to give you some background, Richie

3:12comes in two flavors. There is the light version, which is what we have here, which requires the brains of it to be running on a computer. I have a Mac mini right there that is perfect for running long-running software on, so I'm going to be running my Richie mini software off of this computer, and Richie's going to be communicating over Wi-Fi. There's also the full-on version of the Richie

3:35mini that has a built-in Raspberry Pi. However, I think the light version, if you have a long-running computer, makes a ton of sense, right? So, what did I do? Well, I started off with a fork of the existing conversation app. From there, I used the prompt in the docs to really get started. I then replaced all of the calls to OpenAI with calls to My Hermes, which was really simple because

3:59Hermes has a great API. And this actually gave me a ton of features out of the box considering I use My Hermes for a ton of stuff already. So just simply by connecting it to Hermes, I was able to get memory via Hancho, tool calling of all the tools that my various profiles have set up, coding tasks so it can actually do coding on any of my

4:21machines since they're all connected via SSH. So simply getting those things working was super easy, but the tough stuff here was actually getting the wake word timing correct as well as conversation timing. Basically, the Richie software has voice activity detection. It decides when speech is actually starting and ending based on the audio level that's coming in. If the voice activity detection detects speech or silence, it then starts to transcribe

Around this claim