Era of AI Scalability Begins, Financial Institutions Taking the Lead

Technology Author: Butao Wang Jun 21, 2019 02:01 AM (GMT+8)

In 4Paradigm's "1+N" Strategy, one stands for the ultimate effect that AI could achieve, forming a definite advantage in a company's core business, and N stands for scalability in more scenarios based on unified ideology and platform.

CEO of 4Paradigm spoke at the conference. Photo: Credit to 4Paradigm website

4Paradigm, China's leading artificial intelligence (AI) technology and service provider, unveiled its new line of product SageOne Appliance in a releasing conference on June 20. Focusing on a new paradigm of the industrial revolution, the company exported its “1 + N” strategy based on the newly launched software-hardware integration system.

Founded in 2015, 4Paradigm had finished its Series C round of funding last year, raising over CNY 1 billion with a valuation of around USD 1.2 billion. Among its investors are Sequoia Capital (红杉资本)and Sinovation Ventures (创新工场). It is also the only startup company that was invested in by all Big Five (Bank of China, Agricultural Bank of China, Industrial and Commercial Bank of China, China Construction Bank, Bank of Communications) state-owned commercial banks in China.

Dai Wenyuan (戴文渊), CEO of the company, gave a speech on 4Paradigm’s inception story, AI landing cases as well as visions in the future.

“One stands for the ultimate effect that AI could achieve, forming a definite advantage in a company’s core business,” Dai explained the “1 + N” strategy for enterprise AI transformation. “While N stands for scalability in more application scenarios based on unified ideology and platform.”

After Dai's speech, professionals of 4Paradigm and its cooperating clients further discussed the application landing and the scalability trend in different industries.

Companies in finance, medical research and services, energy, government affairs, media and retail industries have been showing needs in AI application for the purpose of using automation instead of manual, improving customer experiences and cutting costs.

As the forerunner industry in AI transformation and commercial landing, finance took the lead as one of the three sub-forums' theme in the new product releasing conference.

To better illustrate AI transformation in financial institutions, Chai Yifei (柴亦飞), vice president of 4Paradigm explained the current two approaches in achieving scalability. The first approach is to cut in from a single application scenario, then build the platform of full-stop ability after validations of several successful scenarios. This approach is adopted by most of the financial institutions. The second approach is to build the platform ability at the beginning and apply to different scenarios gradually.  Industrial & Commercial Bank of China, which adopted the second approach, has developed more than 20 application scenarios this year by far and is planning to add 120 more scenarios in 2020.

Results are satisfying in banking sector: the recognition rate for anti-fraud in credit card transactions increased 268 times, the success rate for personalized financial products recommendation surged 574%, the accuracy rate of OCR on handwritten bills is 98%, equally matched human performance but in a much shorter time.

Through these practices with banks, 4Paradigm has established a solution covering the full business cycle:

With the ambition to build an AI ecosystem and establish the closed loop of industry recognition, business processes, data, technology and talents, 4Paradigm stressed its core competency in high-dimension algorithm (VC dimension) and real-time response, empowered by its Auto ML and transfer learning technologies. 

At the same time, several challenges are ahead of the road in AI transformation:

1. Lack of AI talents that prevent the faster application of 4Paradigm’s software and hardware;

2. Complicated organizational and recognition issues in scalability;

3. Return on Investment for small companies is not cost-effective due to the consideration of TCO;

4. Lack of AI industrial knowledge and consciousness in accumulating data;

5. Safety and controllability of data in the future.

At the end of the day, similar hesitation and doubts arose when we first saw the coming of mobile Internet or 4G, but the trend was unstoppable. AI is considered to be a strategic technology that leads to a new round of scientific and technological revolution and industrial transformation. AI transformation should not only be the strategic needs of enterprises but also the key to future competitiveness and innovation.

With the ambitious vision of "AI for everyone", 4Paradigm will play an increasingly important role in this exciting field.