Principal Investigator: Prof Lam Kwok Yan
Team: Wong Song Wei
In today’s information age, data is booming in volume and business owners need effective ways of gathering insights from their data quickly. Tau Express aims to provide agencies and businesses with an intelligence management system, which produces an accurate and lean method for ontology development and maintenance. Focusing on core technologies such as Natural Language Processing (NLP), Topic Modelling and Word Embedding, and further combining it with Machine Learning methods, we have built a smart information retrieval engine capable of developing ontologies across knowledge domains and industry sectors. Value created for our users includes a reduction in time needed to train ontologies and gain insights from data analytics, thus leading to smoother workflow for users. Moving forward, we hope to work with potential users across industries, to help unlock the true value of knowledge stored behind their data and empower a leaner yet effective workflow.
Privacy Management Platform
Team Lead: Pandey Avinash
Team: Barun Kumar, Hoang Chu Van
Our privacy management platform target to help organisations, especially SME and Fintech segments to manage key data privacy and protection obligations in line with latest regulations. It’s an easy-to-use compliance management platform, which will automate various underlying processes, providing clear & customised roadmap to address gaps, with real-time tracking. The platform comprises AI driven modular structure, to address diverse requirements and priorities, e.g. Customer & Employee Data. The solution is offered as service, i.e. SaaS model, ensuring platform security is offered as standard feature. For large enterprises, we provide an option of customised on premise standalone solutions as well, with provision to periodic platform updates, with an option to integrate other business reports in the reporting dashboard as a value added service, e.g. management dashboards on targeted marketing. The solution is being received by our potential clients positively, considering the shortage of privacy professionals and the need to automate data privacy efforts based on an organisation’s need to manage heterogeneous and large volume of data in the digital economy.
Principal Investigator: Prof Tony Q.S. Quek
Team: Jay Prakash, Andrei Bytes, Chuadhry Mujeeb Ahmed
Varuthya aims to design usable and secure authentication solutions for web-services and inter-connected devices. As our flagship product, we introduce PhyAuth, a multi-modal multi-factor authentication (M3FA) framework, which would replace contemporary two-factor-authentication (TFA). Our study reveals serious usability and security tradeoffs across industries and methods of TFA. Our solution implements security protocols and frameworks on the pillars of wireless and acoustic sensing, cryptographic primitives, multi-model analysis and design science. PhyAuth has been developed with focus on a human centric security and ensures a closed loop continuous verification of possession and proximity of the token device. While security assurance is higher than before, it expects minimal intersections from humans while authenticating logins. Varuthya’s seamless M3FA, PhyAuth, is expected to be adopted across all sectors which could not adopt TFA either due to the risk of losing user base, cost of deployment or limited services. LLP has helped us identify key market segments, pain points and contextualise our solutions for each of these clients.
Principal Investigator: Assoc Prof Ng Wee Keong
Team: Ong Jenn Bing
Big Data is essential to develop cutting-edge artificial intelligence systems. However, sensitive information can be extracted from big data and therefore may incur privacy issues. We propose a big data shredder that systematically decomposes structured and unstructured data into fragments with partial information. These fragments are non-unique, un-linkable, and not interpretable; they are distributed among multiple devices or clouds with metadata privacy. The increasingly complex digital environments nowadays require simple big data security solutions with distributed trust. Our solution is memory-, bandwidth-, and computationally-efficient; most importantly it is keyless, therefore it overcomes the expensive and complicated key management and distribution problems inherent in current data encryption systems. As a mathematical technique, our solution can be easily combined with existing privacy-preserving technologies to provide layered protection; it can also be easily integrated into existing computing platforms, environments, and processes.
Principal Investigator: Asst Prof Li Yi
Team: Dr Xie Xiaofei, David Berend, Du Xiaoning
Artificial Intelligence has great impact on safety-critical areas, such as autonomous driving, healthcare, public sector and finance. Currently, the technology adoption is facing a common problem: lack of usable data and assuring quality standards. TRAICO Labs is able to verify if an AI system is ready for deployment in the real world. This is achieved by generating the data that tests the critical cases. Thereby, errors in autonomous driving systems can be detected before a crash might happen in the real world and doctors can provide more accurate and confident suggestions to their patients. TRAICO Labs started in the Lean LaunchPad programme with the goal to solve the entire AI landscape: providing an end to end platform, building a community, hosting a cloud environment and consulting the clients. The program and mentorship of Eugene Aseev helped them to identify their core value proposition and challenge it through interviews in the relevant industries.