Anyscale Success Story: Scalable Compute for AI and Python

Since the last few years, the demand for machine learning engineers has exploded significantly. Moreover, IDC estimates that the amount of data created and copied in a year will reach 175 zettabytes by 2025.

The higher demand for expert resources, along with the astronomical rise of data, implies that several companies will be unable to leverage advanced technologies such as machine learning without the right tools and applications at their disposal.

As artificial intelligence usage is increasing and pressurizing computing power, technologists have split the workload from a single computer to thousands of machines, an approach called ‘distributed computing.’ However, building and maintaining distributed infrastructure is complex, requiring at least 5-10 engineers who are experts in distributed computing.

It is when Anyscale comes to the rescue. This AI-based company is the future of the distributed computing industry. It allows developers to easily develop distributed applications, effectively increasing the level of the playing field. Consider reading this article to learn about Anyscale, its startup story, founders, funding, business model, growth, and more.

Anyscale – Company Highlights
  • Company Name – Anyscale
  • Headquarters – San Francisco, California, United States
  • Sector – Software Development
  • Founders – Robert Nishihara, Philipp Moritz, and Ion Stoica
  • Founded In – 2019
  • Revenue – $1B (2021)
  • Website – Anyscale.com

Anyscale - About
  • Anyscale is a fully-managed, enterprise-ready unified compute platform that enables companies to build, deploy, and manage AI and Python applications with the help of Ray. By leveraging the company, developers of any skill level can easily build applications running at any scale, from a laptop to a data center, at lower costs.
  • Located in San Francisco and Berkeley, Anyscale serves developers at industry leaders such as OpenAI, Uber, Microsoft, Intel, Ant Financial, Amazon, and Shopify.

Anyscale - Founders and Team
Robert Nishihara, Ion Stoica, and Philipp Moritz founded Anyscale in 2019.

Robert Nishihara : Robert Nishihara attended Harvard University to complete a BA in Mathematics and the University of California, Berkeley, for a Doctor of Philosophy (Ph.D.) in Computer Science.

He worked as Software Engineer/Developer Intern at Google and JaneStreet; Research Intern at Microsoft Research and Facebook. Later, he co-founded Anyscale and is the company’s CEO.

Ion Stoica : Ion Stoica attended Polytechnic University Bucharest to complete an M.S. in Computer Science and Control Engineering and Carnegie Mellon University to study Ph.D. in Electrical and Computer Engineering.

In addition to co-founding Anyscale, he is also a Co-Founder of Conviva and Databricks. Ion is also a Professor in the EECS Department at UC Berkeley and Co-Director at AMPLab.

Philipp Moritz : Philipp Moritz went to the University of Cambridge to study Master of Advanced Study in Mathematics and University of Berkely for a Doctor of Philosophy (Ph.D.) in Computer Science. Currently, he is the Co-Founder and CTO of Anyscale.

Any scale is a team of approximately 290 employees.

Anyscale - Startup Story
  • Robert Nishihara, Philipp Moritz, and Ion Stoica realized that organizations struggle to get value from Artificial Intelligence. Any scale originated from Ray, which is a free, open-source project undertaken in the UC Berkeley RISELab. This project came from the co-founders’ experience with the scaling challenges around machine learning and AI.
  • Nishihara, Mortiz, and Stoica witnessed strong adoption of Ray in the developer community and came up with the idea of establishing Anyscale- a fully managed platform for Ray. In 2019, the project was first developed in the predecessor AMPLab at UC Berkeley. With the funding announcement, the startup made its product available to the general public for the first time.

Anyscale – Mission and Vision
  • Anyscale’s mission is to enable every developer and team to succeed with AI without worrying about building and managing infrastructure. To make it possible, the company aims to eradicate distributed systems expertise from the critical path of realizing AI’s business potential.

Anyscale – Business Model
  • Anyscale simplifies distributed programming with the power of Ray. Applications developed using Ray can be quickly scaled, which eliminates the need for in-house distributed computing expertise and resources. Any scale is cloud agnostic, and users are allowed to start their application on one cloud and deploy to the other. Moreover, the platform is supported by a rich ecosystem of libraries and applications, which removes the barrier to entry for developing scalable distributed applications.

Anyscale – Products and Services
  • The company offers two products- Anyscale Platform and Ray Open Source, along with multiple solutions, including Data Ingestion, Reinforcement Learning, Model Serving, Ray AIR, Hyperparameter Tuning, Scalable ML Platforms, NLP, and more.

Anyscale – Funding and Investors
  • Anyscale has undertaken 4 funding rounds, raising $259 million. Its latest funding round was raised on August 23, 2022, from Series C Round and collected a total amount of $99 million. The company is backed and advised by 11 leading names in the industry, including Andreessen Horowitz, New Enterprise Associates, Addition, Intel Capital, and many more.

Anyscale – Patents and Trademarks
  • Anyscale is registered with 3 trademarks, the ‘Scientific, Electric Apparatus, and Instrument’ the most popular category.

Anyscale – Growth
  • The estimated annual revenue of Anyscale in 2021 is $45.1 million ($157,867 per employee). In December 2021, the company’s valuation was $1 billion. Furthermore, its employee count grew by 101% last year, and the monthly web visits rose to 55,699 with an 18.62% growth rate.

Anyscale – Partners
Some of Anyscale’s partners are:
  • Microsoft
  • AWS
  • ByteDance
  • Github
  • Slack
  • Discuss

Anyscale – Competitors
The following are some main competitors of Anyscale:
  • IBM Watson Studio
  • Databricks Lakehouse Platform
  • SAS Model Manager
  • InRule
  • Super Annotate
  • MLJAR
  • Gurobi Optimizer
  • Encore
  • Anyscale – Future Plan

With distributed computing growing faster, Anyscale plans to bring Ray to more enterprises that can benefit from its capabilities.