Candy is a voice assistant for makers. An open platform for exploring Artificial Intelligence, Cloud Computing and the Internet of Things.

About the project

Cognitive Candy is a voice-activated virtual assistant for makers and developers. It’s an open platform to build, create and play with:

  • Artificial Intelligence
  • Cloud computing
  • Internet of Things

Why Candy?

I created Candy for a contest as a way for me to learn hands-on some cool cognitive technologies.

During that process, I realized Candy would be a great platform for makers & developers because it would provide them a richer baseline they could use to learn and prototype in their area of interest. It could be creating a scary talking bot for Halloween or maybe experimenting with the latest Natural Language Classifier technology. Candy takes care of the basics.

The goal is to make Candy an easy to use educational tool for kids, but also a powerful open platform for hardcore developers. I want it to be an awesome Developer Experience.

Video for Contest: Demo at the Maker Faire, IBM Blockchain, Real-Time Insights and Watson IoT Dashboard. Overview of Architecture, and UX Design. Length: 10min

How to program it?

The super easy way to program Candy via the “Candy Brain Web Editor”. It supports IFTTT and other simple API integrations. See demo on YouTube.


For more powerful programming, but still easy enough, use Node-RED to create a flow to process the API requests and responses to 3rd party services. An example is coming soon.


Next option is to write your own node.js or Java module, or even a cloud-based app to extend Candy’s logic.

How’s Candy made?


  • Candy Dev Tools includes the Candy Brain Editor and Brain Tap (so you can type instead of speaking to Candy).
  • CandyCloud: cloud application is providing remote management, inbound connections, etc. Stretch goal / under development.
  •  CandyOS is the operating system inside Candy. It manages the user interaction logic and integrations with the outer world. It’s a node.js program written in Node-RED.

Speech Recognition & Synthesis: Candy relies on external Speech libraries. You have options here: you can use free open source libraries like PocketSphinx & eSpeak or you may sign-up for an online service like IBM Watson.

Hardware: Candy is based on the Raspberry Pi. Its enclosure is 3D printed. It has wifi, built-in speaker, microphone, buttons, and LEDs.

A Candy Kit contains a Raspberry Pi, microSD card, power supply, speaker, microphone, button & LEDs circuit board, audio amplifier, enclosure, and mounting hardware.

About the creator

Igor Ramos is the project creator. Find him on LinkedIn and Twitter (@RamosIgorS)

Sign-up to receive Candy updates.


Follow step-by-step instructions and build your own Candy


Extend the functionality of Candy using APIs, IFTTT


Use it day-to-day as your voice assistant!

Hey! Get in touch

Use the form below to contact me and to receive our newsletter with most up-to-date info about the project.

Contact Igor