Gladly, a new era of seamless, safeguarded AI implementation is now closer than ever before. Although a healthy amount of cyber hygiene like downloading antimalware or choosing the most suitable VPN is always encouraged, companies like Nexus FrontierTech will take security questions off your shoulders by deploying a mix of internal and external protection processes.
Today we are speaking with Derrick Liao, Chief Operating Officer at Nexus FrontierTech – an enterprise solutions provider specializing in the development & integration of AI solutions – who will give more insight into the complex world of AI implementation.
Nexus FrontierTech was the result of a collaboration between serial entrepreneur and early-stage investor Danny Goh, and AI Technologists Hajime Hotta and Takaaki Mizuno. As respected academics and practitioners of the field, they were approached for technology consultancy services and software development, starting off their partnership in an organic, client-led fashion, without any sales teams.
With the increasing market demand, they started to specialize in automating data-intensive processes for financial, regulatory, and public sector services, growing the three-person team into over 150 staff across London, Japan, Singapore, and Vietnam.
The journey has been exciting, with the company growing exponentially every year, and seeing success after success in both our technological breakthroughs as well as our client acquisition has been extremely motivational for us.
Our vision is to enable a digital world transformation that maximizes human potential and sustainable value creation – and we’ve already started to see first-hand how our tech has achieved this for our clients!
Our proprietary Al platform, Podder, allows for fast model deployment with an integrated frontend for users in a safe private cloud setting, enabling end-to-end development for organizations to configure, fine-tune, and deploy Al models into any system in an accelerated timeframe. Podder also ensures excellence in Mlops, by handling Testing, Packaging, CI/CD, Diagnostics, and scalability.
Built onto this platform are our core capabilities and a library of 50+ AI models (and growing), specializing in Intelligent Document Processing (IDP), Natural Language Processing (NLP), Computer Vision (CV), and Optical Character Recognition (OCR) to power our solutions. Our AI models are immediately deployable for new users, and can also be finetuned for more bespoke solutions.
We find that there are two main challenges that organizations face when they embark on intelligent automation.
Choice of technology:
Nexus started off by providing bespoke solutions, tailoring each solution to the unique needs of the clients. In our experience, many of our clients failed in process automation because many of the off-the-shelf products available were developed without sufficient input from the experts – a process we call our “X-in-the-loop” – having human expertise to improve the results of digital automation.
The more clients we built our solutions for, the more we managed to configure a truly successful solution, something new to market producing results that are finally good enough for the users.
In our whitepaper on AI Project Management Frameworks and Tools for Success, Hajime shared that while traditional procurement platforms prioritize certainty management, with defined outcomes planned for and tracked against from the very beginning, innovative technology projects require allowances for uncertainty management, determining the broad business goals but embracing a trial-and-error approach along the way.
As artificial intelligence is a new field, organizations need full buy-in from leadership and staff alike to approach digital transformation with an open mind, allowing room for change or new discoveries in their journey. Without this mindset, organizations miss out on opportunities to maximize and scale their digital transformation across every part of their business workflows.
The pandemic drove many to adopt even more digital processes in their daily lives, pressuring financial institutions to improve their digital offerings by tapping on AI solutions to increase their operational efficiency and deliver to client expectations.
Additionally, as banks onboard even more clients via digital channels, they also access a new level of consumer data previously unavailable to them, and they are now figuring out how to use AI to effectively harness the data from these sources for their own business goals.
More currently, the war in Ukraine and the insolvency and debt overhang that followed introduced a new period of high inflation, disrupted supply chains, and slowed growth. With the expectation that we are now facing a period of recession, there are two implications: cost-consciousness in operations and efficiency, and increased risk-consciousness.
These are important developments for Nexus because our main business is helping enterprises save costs in the long run with our suite of risk management solutions that serve as a domain tool for businesses to make risk decisions.
As these sensitive documents cannot be shared externally to create or train AI, but organizations do not have the internal resources for development, the data available effectively becomes dead data. To unlock this value, organizations should consider working with vendors who can create synthetic data for the purposes of building applicable AI solutions, and utilize tools that can effectively anonymize data (e.g. names, personal identifiers, etc) – both of which Nexus FrontierTech has successfully helped our clients with.
Additionally, there is a growing concern for ethics in AI, as traditional AI software has been built with black-box algorithms: where we can only see the input and output, but now how the input is combined. These predictive algorithms are opaque and completely unobservable, creating possible social and ethical ramifications when applied in business-critical decisions.
In order to manage this, we must employ a “White Box AI” approach instead, allowing businesses and tech users to predict the behavior of the AI through its various components, ensuring that the processing of sensitive data via anonymized workflows is still transparent and traceable.
We generally recommend basic cyber hygiene – banning the use of personal computers, providing virtual environments for data handling – and frequent internal operational audits, checking to ensure that our guidelines are clear and training is up-to-date. In more extreme cases, we see organizations that mandate air-gapped devices.
However, organizations also need to understand their own workflows and processes, balancing the need for operational efficiency with data security. In this respect, aspects of their work might require on-premise software, balanced with output that is safe to allow staff access via cloud servers instead.
In AI Implementation, companies like Nexus FrontierTech are very mindful of data security and usually deploy a mix of internal and external processes to safeguard corporate and client data.
The present exploration into web 3.0, decentralized finance, cryptocurrencies, stablecoins, and central bank digital currencies, along with supporting real-time cross-border payments, collaborative data platforms, smart contracts, and tokenization, will likely create upheaval in the financial sector.
Adding more uncertainty to this dynamic stage is how most financial institutions are still in the early adoption stages of artificial intelligence despite keen interest because many organizations have seen spotty success with implementation due to varying degrees of commitment to truly evolve. We expect that as we transition into a cashless society, the financial sector will enter a period of heavy digitization in the coming years, but will face a struggle balancing traditional banking practices with a digital frontier. The financial institutions that can truly commit to implementing and maximizing the operational efficiencies that artificial intelligence and other new technologies offer will be the ones that emerge as the new strongest players in the markets.
With Nexus FrontierTech’s core business in risk management tools that help companies with risk assessment and validation, we have seen how client behavior and demands have driven more innovative and widespread digital solution adoption for better front-end customer service and faster back-end operations.
Since then, however, there remains a lot of room to find innovation in the middle office – business functions like compliance, enterprise risk management, auditing, and so on.
Moving beyond the automation of simple processes like front-end chatbots and back-end execution, the middle office innovations will need to have their validation and verification processes supported by artificial intelligence, ensuring a reduced error rate and more effective policy implementation across functions.
AI has seen a big leap in technological advancement with more minds coming together to work on AI applications. Nexus FrontierTech maintains our leading AI provider position also by staying up-to-date on the latest AI developments, incorporating new techniques quickly into our platform, and using them to improve our service offerings.
In a way, we have to do for ourselves exactly what we do for clients – even as we continue to acquire new clients in our bid to accelerate the digital world transformation!
This article is also published on CyberNews.
1 St Katharine’s Way,
E1W 1UN, London, United Kingdom
36 Robinson Rd, #03-119,
WeWork City House
A03, Mpro Office, 35 Le Van Luong Str,
Hanoi 100000, Vietnam