Sydney-based analytics start up Hyper Anna has received a $1.25 million investment from Reinventure (Westpac’s venture capital arm) and AirTree Ventures.
The company has developed a natural-language-driven data analytics system, which it says can deliver rich insights without the need for a data scientist.
That’s a nice problem to solve. Data scientists are scarce and expensive, with salaries typically coming in north of $120K. For data scientists with social analytics skills, the pay scales often break beyond $200K.
And it is not a problem confined to Australia. There is a global scarcity of data science skills.
The investment will be used to support further development of Hyper Anna’s natural-language-driven data analytics solution, and to help fund a US launch in 2017.
Hyper Anna was set up by data scientists Natalie Nguyen, Sam Zheng and Kent Tian. The founders say they have developed a machine intelligence layer that allows people to interact with any collection of data they wish from any device they choose.
In practical terms, Anna users can ask a natural language question of the data collection using voice, text or email. The tool can be used to review existing data collections and to provide predictive insights without the need for a data scientist.
Hyper Anna is working with IAG, and a project is underway with Westpac.
According to Nguyen, “Our vision is for people to be able to interact with big data by just asking questions in plain English about key drivers of the business like ‘how’s my sales doing’ and the machine intelligence agent just goes away, analyses sales patterns, and brings back important insights for you.”
The solution also leverages machine learning to ensure that results improve with use.
Running on Microsoft’s Azure cloud or on-premise, data collections held in Software as a Service applications such as Salesforce and Google Analytics can be accessed using APIs, while any other data required for analysis can be loaded to the system for interrogation by Anna on an as-needs basis.
Cloud is critical to the outcome, says Nguyen. With no app or software to install, a user just needs Internet access to Hyper Anna with regularly refreshed data.
“The brain of Anna constantly needs to reap feeds of data so it gets smarter — so having a central place for the brain to sit is extremely important and also helps with deployment time. We deploy literally within a day with Azure,” Nguyen says.
“It’s software you don’t see most of the time. There are three ways to interact with Anna, you can either go into a website and basically the data been piped directly to Anna’s brain and sits in an Azure cloud, then you can ask a question in the web site. Or you can cc Anna in an email conversation, so say for example if you email me saying there is a talk at 4pm today about your sales performance and you cc Anna in that conversation, Anna can read from the email context and understand that you will need some information about my sales performance and then email that information. The third way that you can interact is just text a number, just like texting a person, and Hyper Anna will be able to get back to you.”
Because of the machine learning technology that underpins the system, the insights delivered are sculpted to reflect the real intent of the question.
“When humans ask a question we always have an intention for the question. If you have asked me ‘how are my sales doing?’ and I told you that you had sold $100,000 of product, that is actually not very insightful at all. Your intention is to know whether your performance last week was really good or was it not good and why,” says Nguyen.
By mapping the possible intent of a question, monitoring how the response to that question is received and used, and what follow up questions are asked, the machine intelligence agent learns which information has been most useful and ensures that informs future responses.
According to Danny Gilligan from Reinventure, “The Hyper Anna team has built a product which can bring customised analytics services to the large, underserved market of corporate employees and small businesses. Great productivity gains will be possible when front-line decision makers have the customised insights they need delivered to them in real time. One of the cool things about Hyper Anna is it demonstrates the growing maturity of the data value chain, where two clever ex-Quantium data scientists can build a new company to serve specific problems better than anyone else.”
For his part, John Henderson from AirTree described the Hyper Anna team as an exceptional group, with deep technical smarts solving meaningful problems. “Businesses are yearning for easy-to-use data analytics and visualisation tools. Anna is delivering them already, and will only get smarter over time.”
Nguyen said that the investment from Reinventure and AirTree Ventures was important to support the growth ambitions of the company, but also signalled more widely the potential of the technology.