Mycroft Mark II: The Open Voice Assistant

Mycroft Mark II: The Open Voice Assistant is here to change the way we access information, and in a good way. We are moving fast towards a world where AI will determines how easy we access information with our voices alone. AI devices are likely to interact with individuals more than humans interact with each other at home. However, the recent events that occurred between Amazon and Google, where they stopped users from enjoy services accessed through each other’s devices sets a bad precedence. It limits users to certain manufactures who are dominant in the AI market. Such limitation means that users could soon find themselves paying a lot more for services that should have been free to use. The good news is that Mycroft Mark II is here designed to stop that from happening. In this write-up, we are going to look at what is Mycroft Mark II, and why it’s great news for consumers.

Mycroft Mark II: The Open Voice Assistant

The Mycroft Mark II is an Open Voice Assistant that uses open source technology, and relies on the global community of its users to perfect its AI. Unlike Amazon Echo or Google Home who use proprietary technology, and are competing with each other, Mycroft Mark II fully reliant on open source technology and data provided willfully by Mycroft Mark II users. It doesn’t keep any private data from individuals who choose to use it anonymously.


6 Mics and 10-watt Speaker

The Mycroft Mark II incorporates 6 Mics, which work with Acoustic Echo cancellation, Beam-forming, and Noise Reduction to help it locate the speaker’s location and capture their voice without negative interference from background noise. Moreover, its 10-watt speaker design powered by two-inch drivers operating at full range allows it to project a clear and quality sound when responding to the speaker’s queries.

140 Skills

The Mycroft Mark II comes with over 140 skills, with an endless limit. It will continue to learn more as time goes by and it collects more data coming voluntarily from its community of users. The skills vary from the basic ones such as responding to search term queries, creating shopping lists, starting timers, weather updates, and playing music. The advanced responses include telling jokes. Other intermediate skills included as of now include dispensing information from the following sites and services: Roku, npr, twitter, pandora, wink, PHILIPS hue, YouTube, WIKIPEDIA, DuckDuckGo, WolframAlpha, Facebook, and OpenweatherMap.

Design Features

Users can transfer files from and to the Mycroft Mark II using its USB ports, which also allow I/O functionality when making settings such as updating its firmware. User may also use its SD Card slot to make software changes such as updates. In the box, you will also find its start guide for first time users, the Mark II itself, and the power supply.

Other notable features of the Mark II are:

  • Far-field six-microphone array
  • A Xilinx quad-core processor
  • Hardware AEC, noise reduction and beamforming
  • Dual 2″ drivers (10W) with Stereo Sound
  • USB Type A
  • 4 inch IPS LCD touchscreen
  • Wifi and Bluetooth
  • MicroSD card slot
  • 3.5mm Audio Out
  • International adapters for its 18W Power Supply

How It Works

The Mycroft Mark II is programmed to respond to its wake up phrase “Hey Mycroft” which it responds to with a wake phrase to indicate to the speaker that it’s ready for the next instruction. The phrase “Hey Mycroft” is picked up by the inbuilt Mycroft microphone that is ON at all times.

When the next instruction is spoken, the sound is converted into text that can be understood by the machine using its DeepSpeech Speech-to-Text software.

The machine-readable text is then parsed by Padatious Intent Parser to determine what the user intended. To achieve that goal, it analyzes the keyword phrases of the translated text. If the words spoken were, “Set a Timer for Ten Minutes,” the keywords that Mycroft Mark II will interpret are, “Set-Timer-Ten-Minutes.” The timer skill is will be started, and then a response rendered using the Mimic Text-to-Speech software. A typical response that would then be played on the Mark II 10-watt speakers would be, “Okay, ten minutes starting now.”

What Makes MyCroft Standout?

Unlike the Big Tech AI services that gather as much data about as possible, Mycroft’s only interest is to provide you with services. They don’t store data collected from the questions you ask, names of your pets, or other voice snippets it collects from your home. Unless you actively choose to have them include your data in the open data set, they will not do it.

Mycroft Mark II will not analyze your online habits so that it can serve with ads when you use your internet devices. Mycroft respects your privacy and choices. Furthermore, you may customize it to suit your different needs such as wake up phrase, time and weather widgets, as well as friendly robot faces. When it comes to the voices you hear, you may choose the British male voice or continue with the American female voice; Mycroft will include more voices soon. The available wake up phrases you can choose from so far are “Hey Mycroft,” “Hey Victoria,” “Hey Trinity,” and “Hey Indigo.”


Overall, Mycroft Mark II is the Switzerland of voice assistants, a completely neutral player in the AI ecosystem. Mycroft does not have connections with any commercial service provider, which makes them immune for competition wars. That’s why is unique and stands to offer consumers better AI solutions. We hope that you found this article to be informative and insightful.

Similar Posts

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.