Lei Fang, the founder of NineChapters, discovered this pain point in the data industry during his work at Microsoft, and saw the blue ocean of data science platforms. 2011, Lei Fang started to work in the search department of Microsoft Bing, and successively served as a data scientist and a data engineer. He found that more than 800 data engineers within Bing collaborated on a system called Aether, which contained more than 10,000 modules and hundreds of thousands of projects, and the whole team was able to collaborate on the same platform in an orderly manner, with many functions such as management, resource deployment, and improving work efficiency.
In 2013, Fang Lei decided to go back to China to start his own business, and founded Jiuzhang Yunji Technology Company in Zhongguancun, Beijing, hoping to build a collaborative office platform for data science teams to help data scientists and data engineers in the enterprise to better collaborate.
Currently, in most enterprises, the value of data is mainly embodied in superficial data analysis, making data into visual reports including pie charts, line graphs, etc., and then guiding the business. As the data accumulated by enterprises become more and more abundant, the requirements for data analysis become higher and higher, and the past data analysis methods can no longer meet the needs of enterprises.
According to Fang Lei, the value of enterprise data is undergoing a transformation, and data analysis has entered the stage of "augmented analytics", i.e., augmenting data analysis capabilities through machine learning or artificial intelligence. Previously, visual analytics generated insights through visual presentation, but analytics through AI technology generates more powerful insights, such as analyzing hundreds of millions of transactions in financial anti-fraud through models.
In this environment, Nine Chapters Cloud Pole launched the DataCanvas data science platform, hoping to help enterprises apply AI for intelligent transformation.
In order to reduce the difficulty of enterprise application of AI, DataCanvas provides a complete machine learning platform and AI model production platform, which automates a series of difficult data modeling tasks such as data cleaning, feature engineering, and model training. Users do not need to have a professional data science background and programming algorithm capability, but only need to use the AutoML modeling function on the DataCanvas data science platform to complete the massive data processing and data model full life cycle management.
DataCanvas is even more powerful in real-time data processing. For example, in the financial industry, which generates massive amounts of data every day, the need for real-time data decision-making is particularly urgent: determining whether a credit card is stolen at the moment of swiping it, and calculating the best investment solution in the ever-changing stock market ...... all need to be ensured by real-time data processing.
"The core of the nine chapters of the cloud pole is to turn the data into a model, to provide customers with modeling capabilities of the technical tools or services to support all kinds of business scenarios." Fang Lei said.
For the development trend of data science, Fang Lei believes that the barriers of technology are decreasing, and it is more important to popularize the application at this stage. "Lowering the technical threshold is no longer the core problem of the development of data science, the core problem is how to combine the technology with the business in the real scene." Fang Lei said. How to efficiently *** enjoy professional knowledge, how to effectively combine industry experience, business knowledge and data science, artificial intelligence, and ultimately realized in business scenarios, is the enterprise in the wave of artificial intelligence *** with the technical landing problems faced.
In response to this challenge, Fang Lei put forward the concept of "knowledge fusion". "A lot of our human common sense is related to business, machine learning can provide insight into subtle data information in some aspects, but some still need to rely on human skills." Fang Lei said, "To build a model, technology is 30 to 40 percent of it, and the rest is actually business knowledge." He cited a scenario in the financial application: for example, the micro loan model needs to examine the hidden liability risk of the enterprise, the first item to be examined by an experienced audit is whether the borrowing is split to zero, and if the company receives the whole amount of the call from different accounts, there may be a hidden liability risk. This requires human common-sense experience to make judgments, and requires people to turn such experience into a feature inside machine learning, and machine learning can make corresponding risk warnings.
In practice, nine chapters of the cloud extremely enterprise landing AI provides a "four library" solution: through the establishment of "four libraries" - model warehouse, feature warehouse, Scene Template Warehouse and AutoML Recipe Warehouse, to solve the problem of integrating enterprise business knowledge and technical knowledge. After the threshold of data analysis and modeling is lowered, the cost of AI application for enterprises is also lowered accordingly, so that AI can be applied in more business scenarios.
At present, nine chapters of Yunji not only have rich practical experience in the financial industry anti-fraud and precision marketing scenarios, but also continue to land on the innovative application of machine learning in the fields of government, transportation, IoT, real estate, education, etc.
The company has also established its own AI application system, which can be used to improve the quality and efficiency of its products.
In a case of serving the government, NineChapters Cloud Pole cooperated with Qingdao Municipal People's Procuratorate of Shandong Province on the "Case Quality Evaluation System Construction Project". Utilizing the DataCanvas data science platform, intelligent case evaluation is realized through the use of machine learning algorithms, model training, and other technologies, which reduces the workload of manual case handling by 80% and improves efficiency by 80%. In the past, due to the limitations of human resources, only 10%-20% of the cases could be sampled, while 100% of the cases could be evaluated with the help of artificial intelligence technology.
Serving global customers is Fang Lei's next goal. He judged that China's To B business will become mainstream after 3 to 5 years. With the development of cloud computing, the data business of many large companies around the world is on the cloud. "The cloud can become an entrance, and on the cloud we can go to provide global services and participate in competition." Fang Lei believes that ToB business going overseas will definitely go through this process.
In March 2018, nine chapters of cloud pole carried out a B round of nearly 100 million yuan of financing, participated by Red Dot Ventures, Oriental Fortune and other institutions. in January 2017, 50 million yuan of A round of financing was realized.