Unleashing the Potential of Intelligent Automation
Artificial Intelligence (AI) is by no means a novel concept. From HAL 9000 in 2001: A Space Odyssey to iconic Star Wars droids like RD2D and the adorable BB-8, humans have long been dreaming of harnessing the power of AI to fuel technological revolutions.
More recently, AI has been making waves in banking and finance, with widespread applications across different fields. We sat down with Sanjay Uppal, Founder and CEO of Finbots.AI, who has been working with the DBS Institutional Banking and DBS Startup Xchange team on bringing AI technologies into the bank. He shares insights into the potential of AI-powered intelligent automation, and how it has helped redefine business processes in financial institutions and services.
AI in fintech has seen some incredible development in recent years. Tell us more about Finbots.AI and the AI-driven solutions offered.
Finbots.AI was founded in end-2017 to develop cutting-edge AI-driven solutions for financial institutions. After spending over 20 years in a variety of banks across Asia, I realised that AI was the answer to meeting the requirements of the increasingly complex financial services industry. We believe that developing effective AI-driven solutions for financial institutions requires a combination of AI, machine learning and data science capabilities, along with a deep understanding of financial services businesses and technologies.
Our innovation comes from our ability to develop products and solutions that are embedded into our clients’ ecosystems, thus making the transition effective and seamless. Our understanding of financial services has allowed us to reimagine the processes to ensure our clients enjoy the most benefits from AI. These have enabled us to create unstructured (documents, emails, scans etc.) to structured data, credit management, voice-to-text solutions, and advanced and predictive analytics.
So why intelligent automation solutions for finance, risk and operations? What excites you the most about working with such technology?
Financial institutions are facing increasing risks (financial, regulatory, reputational, etc.) and cost pressures due to rising business complexity and scale, need for quicker response, multiple technology platforms, key-man risks and a lack of skills. We believe AI is the solution to many of those challenges and will drive the next wave of transformation in financial services. The advancements in AI in the past decade, and specifically in the last three years, give us a glimpse into the possibilities that lie ahead.
What excites us the most about working with AI technology is continuously stretching the realm of possibilities for financial institutions every day. Furthermore, our understanding of the industry allows us to proactively develop solutions in areas that will best benefit financial institutions.
What are the benefits of using intelligent automation over traditional automation?
Traditional technologies have allowed us to automate complex processes and computations, but with the limitation of only dealing with structured data and defined outcomes. The programmes for running such solutions are largely static and would require re-programming to accommodate any changes or updates.
On the other hand, intelligent automation, which is developed using algorithms based in machine learning technology, works with both structured and unstructured data. As such, the resulting solution will continue to learn in real-time from its experiences, thereby progressively delivering higher quality outputs.
In the context of analytics, traditional automation can provide descriptive analytics and some diagnostic analytics. In contrast, AI enables the development of deep diagnostics, as well as predictive and prescriptive analytics capabilities, which greatly widens the solution’s potential uses and applications.
Can you share more about the technology and solution behind the Investment Product Repository System developed for DBS? How does it address the problem statement and benefit end-users?
We were delighted to work with DBS to address a challenge that cannot be solved using traditional technologies. The bank’s existing processes required working with unstructured data in multiple formats and with inconsistent nomenclature, combining it with internal reference data and extensive manual processing for a business that has grown significantly in scale and complexity. The need was to automate the process, achieve accuracy and faster processing, and enable insights and analytics.
In response, we developed an AI-based application — the Investment Product Repository System (IPRS) — that was not only able to read and extract unstructured data from multiple formats, but also store data in a structured format. It contains an intuitive user-interface with the ability to sort, search and filter data, allowing for ease of analysis and reporting. Our application also allowed users to supervise machine learning, thus ensuring complete accuracy at all times.
Ultimately, the IPRS was able to shorten processing durations from days to mere seconds and has eliminated errors from the process by matching nomenclatures with high degrees of accuracy. It will continue to learn with every iteration of data it processes, while allowing users to maintain control over the learning process.
As AI solutions rise in popularity, what sets Finbots.AI’s IPRS apart from other automated extraction systems? How were you able to use machine learning to create a customised solution?
The IPRS is built using cutting-edge programming and proprietary libraries to deliver a user-friendly application in a short span of time. We have even incorporated a distinct feature that would allow the user to drill down any data from the report to its source, enabling auditability and comprehensive data lineage.
Machine learning algorithms in the application allow for the extraction of specific data from multiple document and file formats. We have also included Natural Language Processing-based algorithms that help facilitate the matching of inconsistent nomenclatures across various source documents and reports, enhancing accuracy and reliability.
Can you describe your experience working with Startup Xchange and DBS’s Institutional Banking Group (IBG) to bring this solution to life?
It was our absolute pleasure and privilege to work with the Startup Xchange and IBG teams! They were vital in the creation of a positive, can-do environment which enabled the success of this initiative. Through working with them, we were able to understand the problem in greater detail, and they were incredibly helpful in responding to any information requests we had. They also helped in providing prompt feedback throughout our development process, spurring our momentum and ensuring we could achieve rapid completion of the project.
Furthermore, the logistical support at DBS Asia X (DAX), the bank’s innovation centre, has been essential in making our development and innovation journey as smooth as possible. The energy at DAX is infectious and we look forward to being back there soon!
Finbots.AI is Singapore-based fintech with key expertise in intelligent automation, data science and machine learning. Under the Startup Xchange programme, Finbots.AI worked with the DBS Institutional Banking Group to develop an AI-powered Investment Product Repository System which enables the capture, storage and extraction of information from institutional investors for analysis.
