Low-maintenance platform with out-of-the-box AI pipelines
Spend more time building AI apps, and less time maintaining systems. Memri Data Wallet offers a serverless, cloud AI architecture with privacy and security at the core. Choose any open source or commercial ML model, connect to Data Wallets via plugin API and run machine learning workflows securely on any data.
impact
Manage less, build more
Low maintenance cloud AI
Instantly connect to Data Wallets and run ML plugins on Memri’s serverless AI cloud platform, saving time and effort on server and security implementation and maintenance.
Privacy by design
Maintain data privacy and security while running AI workflows or interacting with any code or Wallet databases thanks to data encryption at Rest, in Transit and in memory, a Confidential Compute layer, and a permission-centric plugin architecture.
Unique security architecture
With personal encryption keys, only data owners have access and control over the data stored in their Data Wallets, and only they can recover their keys. This way, any system-wide breaches are prevented without the need for additional security measures.
More models, faster implementation
Choose from a growing number of existing plugins with free and paid, open source and commercial models. Or choose and connect a model to your app via our plugin API and start running inference immediately on any data shared by users from their Data Wallets.
Constant access to growing datasets
Eliminate the costs and delays of collecting or acquiring quality datasets, and focus on shipping new and better features. With an uninterrupted access to the ever-growing number datasets permitted by users, you can spend more time building features, and less time searching for compliant data.
Open source and community support
Run Data Wallet anywhere, any time, including locally; there is no lock in. Join our community of app creators and model makers on Discord, take part in discussions or contribute to the development of the core platform.
Start building AI apps in 5 minutes
Install Memri Data Wallet on your chosen server and start building your AI app.
platform
Ready to build, privacy-centric AI architecture
01
Graph database for scalable data interrelation
When users connect new data to their Data Wallet, Memri’s Graph database allows to easily and quickly interlink new datasets and ensures a self-documenting nature of relationships. This structure creates a knowledge base for the user, which is valuable component for building AI applications.
02
Dynamic schema for flexible interoperability between any app or feature
As users connect and share more data from their Data Wallet and your AI product evolves, the dynamic schema allows to easily adjust the structure of data, maintaining interoperability with other apps on the platform, and ensuring uninterrupted access to data in Data Wallets.
03
Inherited security between apps and Data Wallets
App plugins running your code gain only temporary access to (a subset of) the data in each Data Wallet. Plugins can be triggered by an API call from your app, or by the database, without compromising the security of each user’s data. For example, when a new email arrives in an inbox, a spam detection plugin can be triggered to perform machine learning inference just on that email.
04
Interact with Data Wallet using the pyMemri python library
A continually growing pyMemri python library provides an effective interface for interacting with the Memri Data Wallet. The pyMemri python client offers ready scaffolding for plugin authentication flows.
05
App authentication by default with Data Wallet
Users can use their personal Data Wallet encryption keys to immediately sign into and start using your app. If they lose their keys, users can recover their identity independently. You don’t need to set up or maintain user credentials, privacy or security protocols.
Get in touch
use cases
Real life applications
Memri ONE newsreader
Memri ONE is designed to reduce information overload when reading the news. Each user can chose topics, browse the latest news from top sources, read summaries for topics and articles, and chat with an AI about the topic or article to get more context. All interaction and chat data is kept secure and private, but remains open to sharing and ML training.
Simple scalable implementation
Industry use cases
Get ahead with future-forward industry standards in AI.