Google placed a big bet on low-code and no-code software development by launching Vertex AI about a year ago. But with a new release, analysts think that the internet giant may finally be able to make a dent in the highly competitive market.
At the Applied ML Summit on Thursday, Google Cloud announced several new features to Vertex AI, including Training Reduction Server, Tabular Workflow, and Example-Based Explanations, which are aimed at helping customers better utilize machine learning models and cut down on their dependence on trained experts.
“Our performance tests found a 2.5x increase in the number of ML predictions generated through Vertex AI and BigQuery in 2021, and a 25x increase in active customers for Vertex AI Workbench in just the last six months, customers have made clear that managed and integrated ML platforms are crucial to accelerating the deployment of ML in production,” Google said in a blog post.
Google entered the low-code/no-code market in early 2020 with the acquisition of AppSheet, which was already an eight-year-old company at the time of acquisition. Despite the acquisition, Google is yet to be seen as a serious contender in the low-code/no-code market. Analysts believe Vertex could give Google one more shot at making a dent in the audience for low-code/no-code software development.
“Vertex AI with a value proposition of 80% lesser lines of code requirement compared with other platforms to train a model with custom libraries will further enhance Google’s positioning in low code/no code space,” said Pareekh Jain, founder of Pareekh Consulting. “Google is not counted among leading low-code/no-code platforms yet and this will help improve the positioning of Google.”
According to Gartner’s magic quadrant for enterprise low-code applications, leading players in the industry include OutSystems, Mendix, Microsoft, Salesforce, and ServiceNow. Google didn’t feature in any of the four quadrants, according to the report, issued in August last year.
Google has uphill battle in low-code market
Despite players like Oracle, Microsoft, Salesforce and Google offering low-code/no-code solutions, they have not seen the kind of adoption that one would typically expect, given its promise of eliminating coding and allowing people other than data scientists or machine learning professionals to build AI code.
“Low-code/no-code platforms are good for building efficiency and for building simple use cases but often after using them for a while, developers tend to go back to traditional development tools. The challenge is that most traditional LCNC tools come with a huge licensing cost and yet fail to work well the moment you start building any level of complexity in your code,” said Saurabh Agrawal, senior vice president of analytics and CRM at unicorn ecommerce firm Lenskart.com.
“There are three key aspects of any AI project—the data layer, the data visualization layer, and the ML [machine learning] algorithms layer. Most LCNC platforms start work only on one of the layers. Google has strong solutions like BigQuery, Google analytics, and Lookr, which are mainly used in digital use cases. We hope that if the company is able to crack all the layers with Vertex AI in the automation platform approach, it could come out as a strong player in the segment,” Agrawal added.
Low code/no code has opportunity among SMBs
While most vendors are touting low-code/no-code programming as a means to reduce dependency on hard-to-find machine learning talent, analysts believe the bigger opportunity may lie in targeting SMBs that are looking to build simpler solutions.
“So far companies are more focused on the B2B market for low code/no code for attracting business users, but I think the biggest opportunity for low-code/no-code platforms is democratizing tech for SMBs and individuals,” Jain said. “I think Google and Microsoft have better chance for SMBs. It is like the cloud market. It grew because of initial AWS focus on SMBs. Later it became an attractive proposition for enterprises.”
Copyright © 2022 IDG Communications, Inc.