Skytree
Careers at Skytree
Mission
Skytree’s mission is to bring the power of state-of-the-art machine learning to all data scientists, then ultimately to everyone.
History
Martin Hack had been an employee of many startups in the past. At a certain point he began thinking about ideas for his own company. He noticed that there was a lot of buzz about big data, but that there weren’t any definitive products on the market for obtaining insights from this information.
Around this time Hack met up with his old friend Alex Gray, a professor of Machine Learning at the Georgia Institute of Technology. Hack mentioned his plans and Gray mused that a machine learning system (hereafter abbreviated MLS) might be the solution. They observed that there were no commercial MLSs that addressed big data, and that most of them were aligned with specific industries. Their discussion led them to begin work on a company that would develop a new system.
In 2012 the two launched Skytree, with the flagship product Skytree Server - a general-purpose MLS devoted to analyzing big data for businesses. Specifically, the solution identified patterns and outliers, and made predictions and recommendations. The “Sky” portion of the name referenced the fact that the company’s founders had a background in astronomical data analysis (and strong relations with the Astrophysics community), while the “tree” portion referred to the “cosmic tree of knowledge” apparent in numerous mythologies. While the software could be used across sectors, the team built industry-specific solution sets on top of the platform for more customized help.
The firm got off to a great start, with $3.5 million in funding its first year. In 2013 it raised $18 million in a Series A round of funding led by U.S. Venture Partners. Other investors have included Javelin Venture Partners, Osage University Partners, and UPS. That same year the team introduced the “Second Opinion” program, which provided advisory services to help Skytree customers use the system more effectively. Skytree became a success and now has several high-profile customers.
Benefits at Skytree
Business model of Skytree
Customer Segments
Skytree has a niche market business model, with a specialized customer segment. The company’s offering is targeted at enterprises that need to process and analyze complex sets of data.
Value Proposition
Skytree offers four primary value propositions: price, convenience, performance, and brand/status.
The company offers two single-user desktop versions of its software for free: Skytree GUI & Python SDK and Skytree CLI. They are for systems with up to 100 million data elements and are downloadable on its website. It generates revenues through the Enterprise version.
The company offers convenience by making its platform easy to incorporate into customers’ existing infrastructure. Beyond this, it simplifies usage of the system through heavy automation. Its AutoModel technology makes it easy for users to build machine learning models by automating parameter and algorithm selection. Also, its self-documenting models log all datasets, data splits, transformations, algorithms, and results for future access.
The company’s solution utilizes highly scalable, deeply optimized algorithms that enable it to conduct analytics in-memory and enhance performance. Fewer math steps result in what Skytree calls the “fastest machine learning software on the market”. It estimates that the solution quickens machine learning processes by as much as 150 times the speed offered by open source options.
The company has established a strong brand in a relatively short time. It is considered by many to be the leader in machine learning tools for enterprises. It has many prominent corporations as clients, including American Express, eBay, Honda, Panasonic, PayPal, Samsung, Thomson Reuters, and UPS.
Channels
Skytree’s main channel is its website, through which it obtains most customers. The company’s direct sales team also reaches out to enterprises for interest in its products. Skytree further promotes its offering through its social media pages and participation in industry seminars and conferences.
Customer Relationships
Skytree’s customer relationship is primarily of a self-service nature. Customers utilize the service through the main platform while having limited interaction with employees. The company’s website features an extensive “Resources” section that includes “Quick Start” guides, user guides, white papers, case studies, data sheets, analyst reports, videos, and webinars. That said, there is a personal assistance component in the form of phone and e-mail support.
Key Activities
Skytree’s business model entails producing and enhancing its software product for customers.
Key Partners
Skytree operates the Skytree Partner Program, through which it works with organizations to expand and promote its offerings, as well as advocate for machine learning in general. The program has four main types of partnerships:
Alliances – Skytree helps organizations support and sell its software. They have the option of building applications on top of the platform for a more powerful solution. Specific partners include Hudson Data, BrainPad, LG CNS, Carahsoft, Contexti, Sharpe Engineering, and Big Data Partnership.
Technology – Skytree works with top technology firms to ensure its products have interoperability and compatibility with complementary solutions in the analytics world. Specific partners include Cloudera, Hortonworks, MAPR, Amazon Web Services, IBM, Databricks, Tableau Software, Esri Partner Network, and Cisco.
Original Equipment Manufacturers – Skytree works with OEMs to integrate its platform into their offerings. OEMs receive training on the product as part of Skytree Academy as well as opportunities to co-brand and co-market with the company.
Academic – Skytree works with top academic institutions to promote the field of data science as a profession. The company grants the entities usage of its platform, access to its training curriculum, and guest lectures from Skytree industry experts.
Key Resources
Skytree’s main resource is its proprietary software platform. It also depends on many important human resources. Its engineering staff, 92% of whom are Ph.Ds, produces and enhances the platform. It also has a team of academic experts that includes Professors David Patterson, Michael Jordan, and James Demmel of the University of California, Berkeley, as well as Professor Pat Hanrahan of Stanford University. Lastly, as a relatively new start-up it has relied heavily on funding from outside parties, raising $20.5 million from eight investors as of April 2013.
Cost Structure
Skytree has a cost-driven structure, aiming to minimize expenses through significant automation and low-price value propositions. Its biggest cost driver is likely research and development, a fixed cost. Other major drivers are in the areas of sales/marketing and customer support/operations, also fixed costs.
Revenue Streams
Skytree has one revenue stream: the price it charges to purchase its software, which starts at $2,999 and increases based on the amount of data elements involved. Sales staff must be contacted for more detailed pricing information.
Our team
info: Alexander earned a B.S. in Applied Mathematics and Computer Science from the University of California, Berkeley and a Ph.D. in Computer Science from Carnegie Mellon University. He previously served as an Associate Professor of Computing at Georgia Tech.
info: Brad previously served as the Chief Financial Officer of Neighborhood Networks Publishing and the Chief Financial Officer of Monosphere, a predictive analytics software company. He has more than 30 years of experience in finance operations.
info: Robert earned a B.S. in Physics from Colorado State University and an M.S. in Astronautical Engineering from West Coast University. He previously served as VP of Marketing at Apigee and has held leadership roles at SPSS, Oracle, and Moody’s Analytics.
info: Peter previously served as the VP and World Wide General Manager of the BIRT Analytics Group at Actuate. He also held executive leadership roles at Onyx Software and Business Objects. He has 26 years of experience in Business Intelligence and Analytics.
