Silicon Valley startup dotData provides full-cycle data science automation

Date Updated: Tuesday, December 17, 2019

Computer geniuses, especially the ones who produce the coolest, must-have software, are sometimes called rock stars. Maybe because they’re few and far between, everybody admires them and they can be paid an absolute fortune. Among their number, data scientists are perhaps the rarest of rockers—the Mick Jaggers of the highly specialized world of AI and machine learning. Demand for their services far exceeds supply, and according to LinkedIn there’s a shortfall of some 150,000 rock-star data scientists in the US.

Data scientists bring the skills required for successful data science projects. They know how to extract valuable insights from raw data, how to use AI techniques to coordinate and finesse the factors that produce a successful project, from the most complex mathematical manipulations to the grunt work of data acquisition and incorporation. They’re great to have onboard when you can find them, they can add huge value to your organization, and they can be indispensable to your data science projects.

Well, not necessarily indispensable. Not now that Ryohei Fujimaki and his team at dotData have built their data science automation platform, resulting in the company being selected by CRN as one of its Big Data Startups To Know In 2019. Their solution automates the complete data science process from raw business data, through feature engineering, to machine learning. It democratizes and accelerates the end-to-end process from data to production, bringing the fruits of data science to enterprises large and small, shortening project timelines and enabling more people to create successful data science projects.

clip_image002

How does Fujimaki do that? It helps to have been a distinguished data science research fellow. In fact, he was definitely a rock star before putting his own band together and spinning off dotData. The company provides its platform to enterprise clients that include SMBC, Japan Airlines, Seiko Epson, and a host of organizations in the banking, healthcare, and insurance fields.

There are other efforts out there, but most target only the machine-learning process and stop short of providing a complete, end-to-end solution; using what dotData describes as AutoML 2.0, to distinguish it from other AutoML solutions that don't employ AI-based feature engineering. Other solutions may still rely on interdisciplinary and manual efforts to do the toughest, most time-consuming part of the data science process. That hard, specialized work is known as feature engineering. It’s where the real value of all that data is unlocked, creating insight and information out of chaos. And it’s where dotData adds pioneering AI-based technology, completing the process and providing the full-cycle data science process—from data ingestion and preparation, to feature engineering, machine learning, and integration of ML and AI models into production environments. The solution powers the full data science life cycle, dramatically accelerates time to value, and can help shorten project turnaround from months to days.

clip_image004

As a startup, dotData knows the challenges of building awareness and managing longer sales cycles. Visibility is key to getting in front of customers who are looking for solutions. And it’s followed by enabling them to try out and then adopt your products and services easily. That’s why Fujimaki and his team are engaged with the Microsoft for Startups team, experienced professionals who can help launch dotData solutions on Azure Marketplace and provide resources to support marketing and sales efforts. The Marketplace helps take care of global sales and distribution while the dotData team gets going on what they do best—innovating, staying ahead, and growing their pioneering business.

Subscribe to the Microsoft for Startups blog

Subscribe