“HKICT Award 2020 – Fintech Winners’ Sharing Series”- Fintech sharing by FANOLabs
“The Hong Kong ICT Awards” aims at recognizing and promoting outstanding information and communications technology (ICT) inventions and applications, thereby encouraging innovation among Hong Kong’s ICT talents and enterprises in finding creative solutions to meet business and social needs. In particular, the“Hong Kong ICT Awards : FinTech Award”advocates for the innovation of Fintech - one of the lead driving forces in the exploration of new business models for the banking industry.
Organized by the Hong Kong Institute of Bankers, the “HKICT Award 2020 – Fintech Winners’Sharing Series” was held on 11th June 2021. Dr. Miles Wen, the co-founder and CEO of Fano Labs - Winner of Fintech Grand Award and Fintech (Regulatory Technology and Risk Management) Gold Award, was invited to share their award-winning product “Callinter”. Mr. Stephen Leung, Head of IT Wealth and Personal Banking of Hang Seng Bank, Mr. Byron Wong, Technical Director of Asia Private Banking, JPMorgan Chase Group, and Mr Michael Leung, Group Executive Officer of BOA International Financial Group, also attended the sharing session.
While introducing “Callinter” the award-winning AI-enabled speech analytics solution, Miles mentioned that the idea was inspired by the compliance challenges faced by the banking industry nowadays. In addition, he shared a functional demonstration of Callinter, highlighting its ability to accurately recognize speech and showcased its innovative mixed-lingual speech technology features. Miles also shared his view about the needs of advanced AI technologies to improve the efficiency of frontline sales assessment and compliance reviews. During the Q&A session, the panel discussed more about the prospects of Fano Labs, and shared their insights about FinTech developments.
Q: In Hong Kong, a place where multiple cultures collide, how can FANO Labs ensure it has unique advantage while facing international competitors? Miles: In recent years, there have been many companies focusing on AI speech technology, like those in Mainland China and USA, but most of them only target single languages. We noticed the mixed-language usage in Hong Kong and Singapore, and focused on developing multi-language speech recognition. We increased the accuracy rate to over 90%, even in dialogue with poor grammar, heavy accents, or even when speakers mix foreign languages and use various languages in the middle of a sentence or conversation. I think it has given Callinter an unique position on the market.
Q: Is there an estimate of how much human resources, time and money costs can be saved by using AI speech tech? Miles: At this point, we haven’t acquired much data yet. When the product first launched, our primary goal was to gain users’ trust by providing accurate data so they can accept the technology easily. Gradually, when they adapt to this workflow, it can save at least 50% of working time and more than 70% of human resources.
Q: What mentioned earlier is that the use of technology can compress two hours of speech and the machine can listen and analyze it in two minutes. How can this be achieved? Miles: More importantly, AI technologies have the power to scale in ways humans cannot. For instance, one of the most commonly used way for us to “speed up” our AI processing speed is to adopt parallelization. Where we can first split a 2-hour long audio into 120 of short segments, then deploy 120 AI workers to “listen” to different segments of this audio concurrently. Each worker would need about 1 minute to finish “listening” to an audio, but since these workers work simultaneously, the actual real-world time needed to finish analyzing a 2-hour audio is only 1 minute. After that, we deploy another AI worker to piece all these 120 segments together and quickly look through the generated texts and correct any mistakes, which would take 1 minute. So that is how our machines can efficiently use 2 minutes to finish listening to a 2-hour audio.
Q: Final question for Miles, what would be the next step for FANO Labs? Miles: We started Fano Labs by engaging in scientific research. At the implementation stage, we received a lot of help and guidance from our partners and predecessors to put our thoughts into something practical. I am truly grateful for the support we received in this journey. For the next two years, we will focus on the integration of technology and operations. We hope the banking industry will adopt and apply this workflow in their daily businesses in order to increase their efficiency.
Q: We have touched many pain points of banking operation and I’d like to ask Steven, can you share whether you have similar problems in the process of running different types of banks?
Stephen: Indeed, what Miles mentioned are also our daily challenges. As the head of IT Wealth and Personal banking, we always look for trustworthy solution to ensure all records are accurate, complete and clear. It will impose extra pressure to our frontline colleagues if they have to worry about the quality of recordings. Therefore, we do support and look forward to adopting some more advanced and promising technologies, especially in the RegTech area, so that our frontline colleagues can focus on delivering better services to our customers.
Miles also cited two cases to introduce the role of voice data besides regulatory, and its usage in different industries. These cases showed that voice data can not only be used to analyze business bottlenecks, but can also boost sales. Regarding customers’ privacy and data security, Miles stated that Fano Labs is mainly responsible for system upgrades and maintenance. Generally speaking, they do not have access to specific data and information, so to minimize the risk of data breaches.
At the end of the sharing session, all speakers expressed their positive opinions regarding the development of Fintech and RegTech in Hong Kong, and look forward to seeing more derivatives of AI speech technology. In the era of ‘banking everywhere’, the speakers pointed out there will be countless challenges in different aspects. FinTech companies are encouraged to stimulate inspiration and be flexible in problem solving. As one of the most influential local banks in Hong Kong, HASE is looking forward to working with companies that are willing to make breakthroughs like Fano Labs, and inspire more innovators to join us. Together we can dream big and open up a world of opportunities for our customers and investors.