Anyscale is a San Francisco-based AI infrastructure company founded by the creators of Ray, an open-source distributed computing framework developed at UC Berkeley. Its managed platform helps AI developers, researchers and enterprises build, deploy and scale machine learning workloads across cloud and GPU infrastructure. Anyscale supports distributed data processing, model training, fine-tuning, reinforcement learning, batch inference and production model serving. Its platform integrates with AWS, Google Cloud, Microsoft Azure and Kubernetes to simplify infrastructure management and improve compute utilization. Customers include Coinbase, Runway and Canva. In July 2026, Nscale announced an agreement to acquire Anyscale, combining distributed AI software with large-scale computing infrastructure.
Anyscale offers opportunities to solve challenging AI infrastructure problems involving distributed systems, GPU scheduling, large-scale model training, reinforcement learning, multimodal data processing and model serving. Engineers can work with Ray, Python, Kubernetes, PyTorch and vLLM while contributing to an established open-source ecosystem. The company has raised approximately $260 million and serves AI teams at organizations including Coinbase and Runway. Its founders have strong UC Berkeley research backgrounds, while CEO Keerti Melkote brings extensive enterprise technology leadership experience. The careers page reports positive Glassdoor ratings and employee recommendations. Advertised benefits include healthcare, parental leave, flexible time off and learning stipends, varying by location. The proposed acquisition by Nscale could expand infrastructure opportunities, although integration implications remain uncertain.
Programming Language
Programming Language
Distributed Computing
Data Processing
Machine Learning
Model Serving
Machine Learning
Container Orchestration
Container Orchestration
Containerization
DevOps
AI Inference
Cloud
Cloud
Cloud
Co-founder of Anyscale and co-creator of Ray. Holds a mathematics degree from Harvard University and a PhD from UC Berkeley. Helped develop Ray at Berkeley's RISELab and co-founded Anyscale in 2019 to commercialize distributed computing technology. Served as CEO until July 2024.
Co-founder and CTO of Anyscale. Holds a PhD in computer science from UC Berkeley and studied mathematics at the University of Cambridge. Helped create Ray and leads technical development across distributed systems, machine learning and scalable AI infrastructure.
Co-founder and Executive Chairman of Anyscale. Professor of Computer Science at UC Berkeley and co-founder of Databricks and Conviva. Holds a PhD from Carnegie Mellon University. His research focuses on distributed systems, cloud computing and networking.
Co-founder of Anyscale and researcher in machine learning, statistics and artificial intelligence. Distinguished Professor Emeritus at UC Berkeley and research director at Inria and Ecole Normale Superieure in Paris. Previously served as a professor at MIT.
Appointed CEO in July 2024. Previously co-founded Aruba Networks and led its development through its IPO and acquisition by Hewlett Packard Enterprise. Subsequently led Aruba and HPE's Intelligent Edge business. Oversees Anyscale's strategy, enterprise adoption and commercial growth.
Appointed Field CTO in March 2026. Previously worked at Attentive on production AI infrastructure. Advises enterprises on distributed AI architecture, workload migration, training and inference performance.
Leads customer-facing strategy, technical engagement and enterprise adoption at Anyscale. Previously held go-to-market and partnership leadership roles at Confluent and New Relic.
Anyscale
Anyscale and Google Cloud announced native Ray integration with PyTorch on Google Cloud TPUs. The integration supports distributed training, inference and reinforcement learning across GPU and TPU infrastructure.
Read articleAnyscale
Anyscale published an engineering study demonstrating how Ray coordinates robot foundation-model inference, physics simulation and large-scale evaluation for physical AI experimentation.
Read articleAnyscale
Anyscale announced general availability of its Azure-native integration, enabling enterprises to run distributed data processing, model training, evaluation and inference within their Azure subscriptions.
Read articleAnyscale
Anyscale introduced KubeRay Connect in private preview, allowing Kubernetes-based Ray deployments to use Anyscale observability and orchestration capabilities.
Read articleNscale
Nscale announced a definitive agreement to acquire Anyscale, combining its AI cloud infrastructure and GPU capacity with Anyscale's Ray-powered distributed computing platform. The agreement was subject to closing conditions.
Read articleAnyscale
Anyscale announced a public preview of its Azure integration, designed to support foundation-model workloads with enterprise security, governance and infrastructure control.
Read articleAnyscale
Anyscale appointed former Attentive AI infrastructure leader Christian Stano as Field CTO to help enterprise customers deploy production-scale Ray architectures.
Read articleAnyscaleBengaluru, India
AnyscaleSan Francisco, USA
AnyscaleSan Francisco, USA
AnyscaleSan Francisco, USA
AnyscaleNew York, USA
AnyscaleBengaluru, India
AnyscaleSan Francisco, USA
AnyscaleBengaluru, India
AnyscaleLocation not specified
AnyscaleLocation not specified
AnyscaleSan Francisco, USA
AnyscaleSan Francisco, USA
AnyscaleBengaluru, India
AnyscaleSan Francisco, USA
AnyscaleLocation not specified
AnyscaleSan Francisco, USA
AnyscaleSan Francisco, USA
AnyscaleSan Francisco, USA
AnyscaleSan Francisco, USA
AnyscaleSan Francisco, USA
AnyscaleSan Francisco, USA
AnyscaleBengaluru, India
AnyscaleSan Francisco, USA
AnyscaleLocation not specified
