Mamika


THE CUTTING EDGE OF OBJECT

TRACKING BROUGHT TO MARKET

Try the tech demo


What is Mamika?

Mamika is a cutting edge voice assistant designed from the ground up to have a useful purpose at its core. She creates an inventory of all your things, and remembers exactly where they are so that you don't have to. Looking for your keys again? Just say the wake word, ask, and you won't have to waste your time searching.

This core inventory and tracking functionality of Mamika will also be available as an ApplicationProgrammingInterface accessible by existing voice assistants such as Alexa and Google Assistant, making integration even simpler. The goal is to make it as seamless as possible, and to support as many QualityOfLife and smart home functionalities as possible.



Who are we?

We are a small but growing team with a big vision. Mamika has been our dream for years, and our passion project from the very moment we realized it was feasible. Since then, we have worked tirelessly to make it a reality.

Our team is always growing, but the founders are
Gabriele Calvieri and Rodolfo Verde



How does it work?

Mamika leverages the power of LargeLanguageModels and cutting edge Object Tracking to make your life easier. We're working to make this technology efficient and cheap to run, so that anyone can make use of it. We are also developing a new type of sensor that will let AI see the world more clearly.

Learn more about how we'll make it happen



What's next?

While the current version of Mamika, like the tech demo below, is purely a conversational ArtificialIntelligence, we are developing a version capable of object recognition and tracking. This version is in our hearts the "real" Mamika, and it will use a revolutionary new type of sensor, the Spatial Scanner, tailor-made for AI meant to see the world. All visual information will be processed locally, thanks to our proprietary optimizations.

Speech-only version

The preliminary version can already be a good conversation partner and useful by itself thanks to the switch from NLP (like the demo below) to our LLM. It is, however, limited by only knowing what you tell it, thus requiring active usage to have an up-to-date database. Most issues that stem from this are mitigated with stopgaps, and we're always at work to keep adding more functionality and skills. Power users will already be able to get plenty of mileage from Mamika.

Object Recognition

This is what we consider the “real” version of Mamika’s core functionality to be. She will always know where things are, seamlessly and without need for active usage. The possibilities for skills, as well as different devices, that could be improved or even made possible altogether by such a system are endless. The development of this system is where most of our resources will go until it’s no longer just a goal, but a fully fledged part of Mamika.

Try it yourself!

Here are some voice commands you could try. Keep in mind that Mamika understands what you say. You don't need to use any specific command, as long as what you ask makes sense.

Find

You could say "Mamika, find my phone"
or "Mamika, where are my keys?"

Add

This command lets you add new items to a container,
like keys, a phone, a wallet, or anything else.

Try saying "Mamika, add my phone to the desk"

Move

Move your items from one container to another, like "Mamika, move my wallet from the table to the night stand"

Remove

Remove items from containers. For example: "Mamika, remove my keys from the table"

Inventory

Do you want a full list of everything you have? You can say "Mamika, inventory"

...

Location


Schweinfurt, Germany

LinkedIn


Twitter