Sentient.io Launches Pay-per-Use AI and Data Service
- By Paul Mah
- October 30, 2019
Singapore start-up Sentient.io today launched its pay-per-use AI and Data platform designed specifically for developers.
Smart applications can be created by picking from a selection of about 50 microservices, ranging from speech transcription to smart knowledge bases that can be used for chatbots. Various pricing models is available and varies from pay-per-use to monthly subscriptions.
CDOTrends had the opportunity to speak to founder and CEO Christopher Yeo at the launch today.
Data scientists can apply, too
While the various capabilities of the Sentient.io platform were originally built for software developers, its growing data capabilities meant that data scientists can also leverage its data platform to realize greater value.
Yeo explained: “When you are building an AI model, you need to refine and mesh various data layers. For example, by combining our foot fall data with your point of sale (POS) data transactions; you can get insights you might not otherwise gain if you can find the correlation between these two.”
But why wouldn’t programmers or data scientists simply use the AI-centric capabilities available on the top public cloud platforms? Yeo says everything boils down to a lower entry of barrier and a quicker turnaround.
“Many of these algorithms [on the public cloud] still require a lot of work for somebody to be able to understand AI models to tune and refine them. Our own research scientists do make use of these tools,” explained Yeo.
Deployment can be through the cloud, on-premises, or specific data sets as part of the firm's "data-as-a-service" offering. Yeo says Sentient.io partners with telecommunication providers and others to beef up their datasets.
Under the cloud model, services are accessible through REST APIs, which means that data scientists or programmers can leverage their programming language of choice. A web interface is available to manage the various services.
The power of localization
Another strength of the Sentient.io platform, according to Yeo, is its localization. Specifically, its natural language processing (NLP) draws from IMDA’s National Speech Corpus, which means Sentient.io’s AI speech microservices can understand Singaporean accents and speech.
“We worked with IMDA to collect voice samples through their National Speech Corpus, which collected speech samples across different ethnic group and age representation,” said Yeo, who noted that the service has already signed up users such as broadcaster Mediacorp and a big regional bank.
“[These organizations] have tried Siri, Google Home, Alexa for two to three years. While it works for normal type of enquiries. they sometimes get confused. After a long search of many years, they have transited to using our algorithm.”
Yeo acknowledged that growing the language corpus is a “step by step” process with lengthy machine learning training. The next country on his language list? Indonesia.
What’s next
So what’s next for Sentient.io? Ultimately, Yeo sees his company as offering an answer to the shortage of AI specialists around the world. It opens the possibilities of AI to any developer with coding knowledge, regardless of their background in AI.
“We have a goal of benefitting one billion people with the use of AI. We believe that if we can use AI-as-a-Service platform to reach out to one million developers, that they amplify it by each reaching out to another one million users of their apps.”
Yeo also wants to broaden his company’s offerings. “The whole concept of scaling our platform internationally is anchored on the belief that if you put synergistic microservices together – and a variety of them, you can actually build very sophisticated solutions.”
To achieve that, Sentient.io is committed to build algorithms themselves, or where necessary, to license them from universities or other organizations to create a rich AI and data-centric ecosystem.
“AI is a leveler of both intelligence and capabilities and also levels up wealth creation. Our vision in the world today is benefitting humanity with augmented intelligence. We believe that everybody should benefit from AI,” Yeo summed up.
Paul Mah
Paul Mah is the editor of DSAITrends, where he report on the latest developments in data science and AI. A former system administrator, programmer, and IT lecturer, he enjoys writing both code and prose.