Amazon Web Services is working on a cloud-based artificial intelligence program code-named “Ironman” that it intends to officially announce at the latest iteration of AWS re:Invent 2017 set to start on November 27th in Las Vegas, Nevada, The Information reported on Monday, citing sources familiar with the company’s plans. The tech giant’s new initiative is supposedly meant to entail a “data warehouse” service different to the cloud solutions already offered by the company, albeit largely designed for the same purpose – centralizing data with the goal of facilitating its discovery via end-user queries. Ironman is said to be specifically targeted at machine learning and general artificial intelligence companies seeking a high-performance cloud service capable of powering their demanding creations.
Machine learning automation startup DataRobot is supposedly one of the firms working with Amazon on Ironman, as is Domino Data Lab, a San Francisco-based company developing machine learning collaboration solutions for data scientists. It’s currently unclear how involved Amazon’s smaller partners are in the initiative, though the majority of them is likely set to use Ironman going forward and will also help the Seattle, Washington-based company turn the program into a go-to solution within certain niches, insiders suggested. Ironman will reportedly be Amazon’s most capable solution for processing large amounts of machine learning-related data to date, thus being positioned to directly take on Google Cloud and Microsoft Azure, two alternatives that are presently said to be more versatile when it comes to fulfilling the needs of contemporary AI startups.
AWS re:Invent 2017 will also reportedly see the announcement of a new cloud-based deep learning service meant to replace Amazon’s existing offering in this segment and compete with Google’s TensorFlow. The solution is understood to be highly automated in nature and has been in development for over a year now, with another rumored upgrade over its predecessor being the ability to apply it to unlimited data sets. The two upcoming platforms should mark the latest step in Amazon’s efforts to promote its cloud offerings as viable solutions for AI companies, thus building on their existing position in the commercial segment. AWS re:Invent 2017 will be a five-day conference set to be concluded on Friday, December 1st.