- Former Amazon Web Services and Google engineer Bindu Reddy wants to make artificial intelligence more accessible for small companies.
- Reddy has founded a startup, called RealityEngines.ai, with her former colleagues from Google that makes a cloud based software tool to help companies make their own AI models for the workplace tools they are using.
- Even before the company had a product, it raised a seed round of funding led by former Google CEO and Chairman Eric Schmidt and Google founding board member Ram Shriram in July 2019.
- The startup has now officially launched with a tool that had created templates for common AI uses in a company, that gives companies a place to start developing an AI model.
- Click here to read more BI Prime stories.
Bindu Reddy wants to give more people access to artificial intelligence. The former Amazon Web Services and Google engineer, wants small and medium sized companies to be able to use AI in the same way that large companies do, and make sure big tech companies aren't dominating the sector.
Reddy previously started the AI verticals division at AWS, creating AI for particular domains or use-cases. Earlier in her career she was the Head of Product for Google social apps, where she helped build Google+, Blogger Google Video, Google Docs and Google Sites.
She said during her time at Google and Amazon Web Services she noticed that big companies had a gap between the research being done in AI and the products being developed. She realized that a startup could be more nimble and forge a deeper connection between research and product development.
She decided to start a company to address that need. She founded the startup, called RealityEngines.ai, with two of her former colleagues from Google, and it makes cloud based software to help companies make their own AI models for the workplace tools they are using.
Even before the company had a product, it raised a $5.25 million seed round of funding led by former Google CEO and Chairman Eric Schmidt and Google founding board member Ram Shriram in July 2019. Other investors included Khosla Ventures. A few months later, the startup has officially launched with products that help companies incorporate AI into their existing workplace apps.
"How do we make sure that medium sized companies who don't, or even large companies who don't have large ML teams or data science teams and who struggled with kind of deploying, cutting edge deep learning, how do we ensure that they can deploy deep learning models very effortlessly and easily?" was her thinking behind the company Reddy told Business Insider.
She said she saw a lot of companies struggling with how to use and develop AI. Hiring engineers to build it from the ground up is too expensive for some companies and even if they are able to find talent, they don't have enough data to use AI.
"Even though there are lots of cloud tools, a lot of them struggle to work with those cloud tools. So we thought it'd be easier to give them a package service where they just literally have to use a simple API to plug and play into their business app or process and they can get good AI predictions," Reddy said.
Reddy and her team have taken a template approach, identifying many common uses for AI in the workplace and creating templates for users. Those uses range from financial planning, sales and marketing, to fraud and security, she said.
Reddy said the main issue that smaller companies run into when trying to use AI, is that they often don't have enough data to make the tools. To develop AI, you need a lot of data, which is why companies like Google, Facebook and Amazon have had success with it, they are collecting tons of user data that they are using to train their AI models.
Reddy and her team are solving that by giving users the ability to create "synthetic data" that looks like the data they're collected and can be added to it to create an AI model.
"It's kind of like I have, only 20 examples in the past of forecasting, and typically these AI models would require you to get thousands of examples. So what we're doing is, we're taking your original data sets, creating synthetic data, then adding both up, and then making the model," Reddy explained.
While using synthetic data may not make sense to those not familiar with AI, Reddy explains that the synthetic data is very similar to the actual data collected for that AI model, and having that manufactured data helps the AI itself become more accurate.
Reddy says right now RealityEngines.ai has 17 customers and are now looking to grow their customer base.
Got a tip? Contact this reporter via email at pzaveri@businessinsider.com or Signal at 925-364-4258. (PR pitches by email only, please.) You can also contact Business Insider securely via SecureDrop.
Get the latest Google stock price here.