Build your next authentication system securely on AI and not databases!
The Toridion authentication API service uses deep learning algorithms to store, analyse and validate credentials and sensitive identifying information in a highly secure and efficient probabilistic neural network rather than traditional lookup tables and databases.
In addition to the use of industry standard secure one way hash encryption, authID API employs Quantum Random Number Generator (QRNG) on platform to further encode all data passing through the Neural Network
The database is dead
A Review of Artificial Neural Network for Secure Access Authorization indicates neural networks offer a faster and more secure solution than traditional database approaches because when using tables, an intruder may be able to read or alter passwords. Therefore storing information in password/credential tables may be a potential threat to the security of the network. According to Dahiya Password-based authentication mechanisms are used broadly to identify and authorize users, because the method is cheap, easy and 'quite accurate'. The conclusion of Gupta and Khodke's seminal paper concludes “The use of ANN for secure Access authorization provides the benefit to eliminate the disadvantages of maintaining the conventional verification/password table and this makes the system much more secure”. The overwhelming evidence is that machine learning approaches to credential storage and validation are the future of information security. However, until now the lack of robust and well defined ANN implementations has led to a surprisingly low take up.
Machine Learning – The future of information security
Toridion's authentication API offers a best in class solution built upon state of the art probabilistic artificial neural networks. A set of built-in classes provides an accelerated path to m.v.p reducing cost and technical debt. The future of information security is here now.
Immediate benefits include:
- Massively parallel search of >= 90,000,000 searches per second
- Near instant training
- Add, amend, lookup and validate credentials in split second timescales
- First line security with double encrypted pre-presented credentials
- Second line security provided by fully connected NN layer storage
- Super low memory footprint requiring extremely low RAM and no lookup tables
- Simple to use secure API interface PHP and JS/JSON ideal for mobile integration
- SDK for rapid integration into new and existing developments
- Virtually un-hackable enterprise grade security
- Share and validate biometric & personal data anonymously
- On site, on cloud or hybrid deployment (multi site distribution option)
TORIDION TQNN Artificial Neural Networks Typical applications
This is not a database. TQNN is a realtime smart data fabric allowing the operator to build layered applications across a vast number of unconnected endpoints.
- Track/Trace Tourist Visas in realtime
- Border control checks
- Child protection systems
- National Health systems
- National scale mobile phone MAC address tracking
- Residency conditions enforcement
- On street person identification/validation
- Smart CCTV networks
- Multi credential authentication for banking and military applications
- Instant watchlist alerting
To fully realise the potential power of captured data, organisations and governments need to be able to access it. But access alone is not enough. To build a truly smart nation, governments need to provide data systems that allow external agencies to interrogate data securely in a controlled way without allowing information to fall into the wrong hands. Smart infrastructure built upon Toridion TQNN systems offer unprecedented capability and ease of integration whilst building a system fit for the future.
APi access will be available in summer 2019 to selected partners as part of a beta testing phase. If your organistaion would like to discuss using Neural Network based authentication or explore the possibilites we suggest you contact us directly to discuss a technical engagement.
 MS. Khushboo R. Gupta , Prof. P. A. Khodke : A Review of Artificial Neural Network for Secure Access Authorization [International Journal of Engineering Research & Technology ]
 Shahbaz Reyhani Mehregan Mahdavi : User Authentication Using Neural Network in Smart Home Networks [International Journal of Smart Home]
 Menal Dahiya Department of Computer Science, Maharaja Surajmal Institute, New Delhi : User Authentication Mechanism Based on Neural Networks [IJCSMC, Vol. 5, Issue. 5, May 2016, pg.563 – 566 ]