Intelligence Qubic: Empowering Future with Evolutionary AI
While the APAC region has lagged behind North America or Europe in terms of technology adoption, AI is one area where this market is expected to lead the growth globally. According to IDC, organizations in the APAC region are expected to spend $6.2 billion on AI systems in 2019, which is indicative of a 64 percent growth as compared to the previous year. From providing better customer service to enhancing operational efficiency, AI can create dramatic impacts for organizations. AI offers genuine, tangible potential for driving the next phase of economic growth in the APAC region.
Despite these benefits, only a minority of companies are leveraging AI today. Many businesses are still far away from making the best use of this technology. This is mainly because the expertise in AI is limited to a relatively small group of top data scientists. In other words, AI is not democratic enough, making it difficult for organizations to find the right and adequate resources to deploy AI. It is quite hard and cost-prohibitive for organizations to recruit or train enough numbers of data scientists to tackle with various kinds of AI related solution requirement. Culture conflict is another hurdle in adopting AI for organizations. Since AI is exaggerated and is perceived to ‘replace’ humans in the future, employees are resistant to adopt AI, fearing job loss, which is not valid.
To remove these barriers and unleash the power of AI, Intelligence Qubic (iQubic), an AI technology and solution provider positions itself as best AI partner for Enterprises to maximize the value of AI for operational efficiency and profitability improvement. The company has developed its core product—‘DarwinML’, which is known as an automated machine learning (AutoML) platform.
DarwinML aims to automate the process and remove the skill barrier of applying machine learning. Using it, junior business analysts can easily develop machine learning models within a short period and solve complicated business requirement, instead of spending time to annoying tools and confusing ML concepts. iQubic is inviting Enterprise clients to test DarwinML online.
By DarwinML, iQubic’s vision is to make AI easy to access and be adopted for necessary business lines in enterprises. In addition to the similar functions as other AutoML tools, the differentiated value for DarwinML is it can generate machine learning models better than human designed with the naturally powerful evolution algorithm to unleash the model designing ability. “Moreover, it also tries to empower the explanatory feature to make machine learning not black box anymore,” says Gary Shih, CEO, iQubic.
Despite these benefits, only a minority of companies are leveraging AI today. Many businesses are still far away from making the best use of this technology. This is mainly because the expertise in AI is limited to a relatively small group of top data scientists. In other words, AI is not democratic enough, making it difficult for organizations to find the right and adequate resources to deploy AI. It is quite hard and cost-prohibitive for organizations to recruit or train enough numbers of data scientists to tackle with various kinds of AI related solution requirement. Culture conflict is another hurdle in adopting AI for organizations. Since AI is exaggerated and is perceived to ‘replace’ humans in the future, employees are resistant to adopt AI, fearing job loss, which is not valid.
To remove these barriers and unleash the power of AI, Intelligence Qubic (iQubic), an AI technology and solution provider positions itself as best AI partner for Enterprises to maximize the value of AI for operational efficiency and profitability improvement. The company has developed its core product—‘DarwinML’, which is known as an automated machine learning (AutoML) platform.
DarwinML aims to automate the process and remove the skill barrier of applying machine learning. Using it, junior business analysts can easily develop machine learning models within a short period and solve complicated business requirement, instead of spending time to annoying tools and confusing ML concepts. iQubic is inviting Enterprise clients to test DarwinML online.
By DarwinML, iQubic’s vision is to make AI easy to access and be adopted for necessary business lines in enterprises. In addition to the similar functions as other AutoML tools, the differentiated value for DarwinML is it can generate machine learning models better than human designed with the naturally powerful evolution algorithm to unleash the model designing ability. “Moreover, it also tries to empower the explanatory feature to make machine learning not black box anymore,” says Gary Shih, CEO, iQubic.
The Core of DarwinML
Apart from AutoML for model design process simplification, the design concept of DarwinML is missioned to generated models better than humans based on evolutionary algorithm (EA), which is a generic population-based metaheuristic optimization algorithm as its core engine. The EA powered DarwinML platform allows clients to automatically generate the ‘best’ fitting models from scratch with no reference models in place based on provided data and production environment. To enhance EA for efficiency purpose, DarwinML applies several technologies to ensure that it can generate ‘best fit enough’ models within endurable time and computing power and fit for various kinds of measurement. Most importantly, it can discover sophisticated models from all necessary angles to ensure business target is met with minimal participation of AI experts.
With the DarwinML’s core capability, iQubic has developed DarwinML product family including data fusion for data processing, AI studio for model generation, and inference platform for product system connection via API. Clients can use DarwinML platform to design models in an affordable manner within days even on a simple workstation and deal with regularly changing business data and customer behaviour by refining the production models based on its workflow process design, which is to meet the requirement for AI deployment. “Developing a client-oriented solution focusing on AutoML technology is our key to create a unique value proposition in the market,” says Jerry Qian, CTO, iQubic.
DarwinML focuses on three areas of AI—machine learning, time series, and image processing with miniaturization now and plan to extend to other areas such as recommendation system and universal table-text recognition in soon coming future.
Industry Solution Oriented with Customer Testimony
With DarwinML, iQubic developed several industry-oriented solutions by working with clients or partners.
For instance, ‘DarwinML Fintech solution’, which is developed for business analysts in finance and insurance industry to generate machine learning models for risk management related solution, such as anti fraud, scorecard and finance insurance policy reservations with a friendly GUI portal, business analysts can create AI model by merely clicking the mouse less than ten times.“The value DarwinML Fintech provides is an end-to-end modeling solution, from data preparation to inference running. The models generated by DarwinML are proven with higher accuracy and performance compared to humans,” says Shih. Moreover, a creative new module ‘Rule generators’ will be offered soon to provide recommended rules set for business analysis. This will improve the existing rules capability and enhance the explanatory of machine learning models, which is critical for anti-fraud and scorecard for credit rating solution in finance institutes.
Apart from AutoML for model design process simplification, the design concept of DarwinML is missioned to generated models better than humans based on evolutionary algorithm (EA), which is a generic population-based metaheuristic optimization algorithm as its core engine. The EA powered DarwinML platform allows clients to automatically generate the ‘best’ fitting models from scratch with no reference models in place based on provided data and production environment. To enhance EA for efficiency purpose, DarwinML applies several technologies to ensure that it can generate ‘best fit enough’ models within endurable time and computing power and fit for various kinds of measurement. Most importantly, it can discover sophisticated models from all necessary angles to ensure business target is met with minimal participation of AI experts.
With the DarwinML’s core capability, iQubic has developed DarwinML product family including data fusion for data processing, AI studio for model generation, and inference platform for product system connection via API. Clients can use DarwinML platform to design models in an affordable manner within days even on a simple workstation and deal with regularly changing business data and customer behaviour by refining the production models based on its workflow process design, which is to meet the requirement for AI deployment. “Developing a client-oriented solution focusing on AutoML technology is our key to create a unique value proposition in the market,” says Jerry Qian, CTO, iQubic.
DarwinML focuses on three areas of AI—machine learning, time series, and image processing with miniaturization now and plan to extend to other areas such as recommendation system and universal table-text recognition in soon coming future.
Industry Solution Oriented with Customer Testimony
With DarwinML, iQubic developed several industry-oriented solutions by working with clients or partners.
For instance, ‘DarwinML Fintech solution’, which is developed for business analysts in finance and insurance industry to generate machine learning models for risk management related solution, such as anti fraud, scorecard and finance insurance policy reservations with a friendly GUI portal, business analysts can create AI model by merely clicking the mouse less than ten times.“The value DarwinML Fintech provides is an end-to-end modeling solution, from data preparation to inference running. The models generated by DarwinML are proven with higher accuracy and performance compared to humans,” says Shih. Moreover, a creative new module ‘Rule generators’ will be offered soon to provide recommended rules set for business analysis. This will improve the existing rules capability and enhance the explanatory of machine learning models, which is critical for anti-fraud and scorecard for credit rating solution in finance institutes.
A testimony to DarwinML Fintech is Yeepay, one of the top payment companies in China. The client used machine learning models designed by DarwinML for their anti-fraud business service increasing the accuracy rate from 80 percent to 95 percent. Similarly, DarwinML helped another client called Vbill finance. In addition to anti-fraud, Vbill finance uses DarwinML to create models for loan application credit rating and target to improve 25 percent at least for the first overdue rate. With the success, Vbill now plans to extend to DarwinML’s capability for improving other solution development, such as loan product design, rejection inference and more.
Another case from iQubic is working with a medical solution provider to offer Alzheimer’s disease (AD) pre-screening and treatment services. DarwinML provides AI model design capability to assist in each stage of AD pre-screening and treatment service, from online pre-screening, blood analytics, MRI image classification, and to recovery training policy recommendation services. For iQubic’s partner, that designing AI models now is an easy task for staff, as the process of data collection and preparation is crystal enough and less time-consuming. With DarwinML, the vendor also has chance to design multi-task and multi-stage business models.
"Developing a client-oriented solution focusing on automl technology is our key to create a unique value proposition in the market"
“iQubic is working with several enterprises and partners now by providing DarwinML platform as well as professional services to equip them with skills to leverage AI to create business value in different manners” says Qian.
Innovation on the Cards
Along the innovation front, iQubic consistently tracks and pursues technology advancements to deliver business value to its clients. The company is backed by a hardworking team of experts having rich expertise and knowledge in the AI space geared toward bringing new enhancement to the platform. In addition to regular business target incentive to drive for performance, the company also focuses on top talent and leader identification and development. Identified leaders are invited to join the core business and technology board to share their views and recommendations to channelize the company direction, set goals, and lead the team.
From the technology perspective, it is essential for DarwinML to continue building an AI model from scratch with complete automation capability that is way advanced as compared to human-generated model. Here, the critical factor is the convenience to deal with data preparation. Keeping this in mind, iQubic plans to launch DarwinML 2.0 GA in early next year, which will include a SaaS-based offering with hybrid cloud to allow more enterprises to access and use DarwinML easily. The solution will be introduced in China and will then extend to Asia pacific. The company will also continue working with core partners to provide more DarwinML inside industry and technology solutions.
Another case from iQubic is working with a medical solution provider to offer Alzheimer’s disease (AD) pre-screening and treatment services. DarwinML provides AI model design capability to assist in each stage of AD pre-screening and treatment service, from online pre-screening, blood analytics, MRI image classification, and to recovery training policy recommendation services. For iQubic’s partner, that designing AI models now is an easy task for staff, as the process of data collection and preparation is crystal enough and less time-consuming. With DarwinML, the vendor also has chance to design multi-task and multi-stage business models.
"Developing a client-oriented solution focusing on automl technology is our key to create a unique value proposition in the market"
“iQubic is working with several enterprises and partners now by providing DarwinML platform as well as professional services to equip them with skills to leverage AI to create business value in different manners” says Qian.
Innovation on the Cards
Along the innovation front, iQubic consistently tracks and pursues technology advancements to deliver business value to its clients. The company is backed by a hardworking team of experts having rich expertise and knowledge in the AI space geared toward bringing new enhancement to the platform. In addition to regular business target incentive to drive for performance, the company also focuses on top talent and leader identification and development. Identified leaders are invited to join the core business and technology board to share their views and recommendations to channelize the company direction, set goals, and lead the team.
From the technology perspective, it is essential for DarwinML to continue building an AI model from scratch with complete automation capability that is way advanced as compared to human-generated model. Here, the critical factor is the convenience to deal with data preparation. Keeping this in mind, iQubic plans to launch DarwinML 2.0 GA in early next year, which will include a SaaS-based offering with hybrid cloud to allow more enterprises to access and use DarwinML easily. The solution will be introduced in China and will then extend to Asia pacific. The company will also continue working with core partners to provide more DarwinML inside industry and technology solutions.
December 17, 2019
