Amazon Web Services wants to make it easier for people to get the most out of machine learning in the cloud, and it just unveiled a new consulting program that’s aimed at helping folks get off the ground.
Called the Amazon ML Solutions Lab, the program will provide customers with access to machine learning experts from Amazon who can help them tackle business problems using purpose-built intelligent models. It’s designed to help businesses without extensive machine learning expertise get their problems solved using the latest systems, tailored to tackle particular problems.
While machine learning systems built with the latest techniques can address issues computers were previously unsuited for, building them requires extensive expertise. People with the sort of knowledge and talent necessary are in high demand, so it can be hard for companies to build ML teams in-house.
What’s more, it can be difficult to bring intelligent models built with machine learning from testing into production so that they can actually deliver the results they were built for. Amazon is aiming to solve both of those problems with this lab.
Amazon expects typical ML lab engagements to take between three and six months, depending on the task at hand. During that time, the company will offer developers courses on the concepts behind machine learning, followed by weekly check-ins about the progress of a project.
In some cases, the company will also send members of its team to work with customers for extended periods of time. Companies will also receive materials for additional training, guidance about change management for ML, and advice on establishing centers of excellence for ML within their organizations.
Companies that already have high-quality, clean data that’s set up for machine learning with the correct labels can sign up for the Amazon ML Solutions Lab Express program. It’s an abbreviated experience that starts with a one-week intensive boot camp on Amazon’s campus, followed by three to four weeks of guidance as a customer moves to train and implement a model.
AWS doesn’t guarantee that customers will get a working model that solves all their problems. While machine learning is good at tackling a lot of things, the field hasn’t solved every problem under the sun, and some don’t have clear-cut solutions already developed.
There’s also the lingering question of this program’s cost. Amazon unsurprisingly hasn’t posted a price to its website, and a representative for the company didn’t respond to a request for comment by press time.
The lab already has customers lined up including The Washington Post (which is owned by Amazon CEO Jeff Bezos), the Toyota Research Institute, and Janssen, the pharmaceutical arm of Johnson & Johnson.
It’s a move that makes sense for Amazon, since helping customers get machine learning applications off the ground can drive adoption of its cloud services.
Building out a consulting arm that can help businesses implement machine learning in their products isn’t exactly a new approach among cloud providers. IBM has a massive consulting business that it uses to assist with deployments of its Watson technology, while Microsoft operates a set of IoT and AI Insider Labs designed to help customers get projects going. Startups like Element AI and CrowdFlower also have programs for businesses implementing machine learning, as do major consulting companies like Accenture and Deloitte.
The news comes a week before AWS re:Invent, the cloud provider’s massive conference for customers, held in Las Vegas. AWS is expected to announce a massive raft of new features and products at the event, in addition to promoting its recent releases.