News Releases

SAN FRANCISCO, Sept. 16, 2024 /PRNewswire/ -- Outerbounds, the developer of Metaflow, a leading open-source ML/AI framework co-created with Netflix, today unveiled key platform enhancements. The new features support the full lifecycle of ML/AI projects, enabling enterprises to:

• Develop ML/AI systems faster and securely within their own cloud environments using cutting-edge development tools. The new features include built-in solutions for software supply chain management, allowing container images to be baked in seconds, drastically simplifying and speeding up ML/AI experimentation and production deployments.
• Scale the training of even the most demanding models across GPU clusters with robust fault-tolerance.
• Deploy highly available production workflows on Outerbounds' managed platform, including state of the art GenAI models, ensuring SLA guarantees and the security required by regulated industries and innovative startups.

The platform manages a wide range of use cases, from supervised machine learning tasks like fraud detection and medical image analysis to cutting-edge generative AI applications, including private large language models (LLM), powered by state-of-the-art models from NVIDIA NIM microservices, seamlessly integrated into the platform.

Savin Goyal, CTO of Outerbounds, noted, "With open-source Metaflow, which we launched at Netflix five years ago, we're uniquely positioned to collaborate with hundreds of leading ML/AI organizations like Amazon, Ramp, Autodesk, and Mozilla - and of course Netflix. Together, we've developed the features they need to power real-world ML/AI use cases today. These features are now available to Outerbounds customers, enabling them to develop and deliver projects significantly faster, giving them a competitive edge in the race to drive sophistication and efficiency across industries."

Powering ML and AI across industries

Several companies have already seen transformative results using Outerbounds:

BlackCrow AI, a marketing analytics firm, cut its ML model deployment time from nine months to just two months. Camelia Hssaine, Head of Data Science, said: "Outerbounds made deployment approachable, and the ability to track experiments has been a game-changer."

Setpoint, a fintech startup optimizing asset-backed lending, increased its compute capacity significantly. Russell Brooks, Head of Machine Learning Engineering, shared: "Outerbounds allows us to optimize financial problems at a scale we couldn't have imagined before, completely eliminating our infrastructure burden."

Tala, a global microlender, increased model deployment velocity by 500%. Will High, Head of ML/AI Data Science & Engineering, noted: "With Outerbounds and NVIDIA NIM, model retraining is automated, and onboarding data scientists now takes days instead of months."

MSAID, a proteomics company focused on drug discovery, accelerated its data processing tenfold. Siegfried Gessulat, Technical Lead for Machine Learning, said: "Outerbounds freed us to focus on research, not ML operations. Our pipelines now run 10 times faster."

Learn more

To learn more about the new and upcoming features, visit the announcement blog post. Outerbounds is a proud sponsor of the PyTorch Conference, taking place on September 18-19 in San Francisco, where attendees will have the opportunity to experience the new platform features in action.

About Outerbounds
Building on open-source Metaflow and spinned off from Netflix, Outerbounds was founded in 2021 to provide a fully managed platform for AI/ML systems for teams to design and develop applications and deploy them on a scalable, enterprise-grade platform. The company has helped hundreds of organizations — from startups to highly regulated banks — deploy AI, ML and data apps from prototype to production, faster. Learn more at Outerbounds.com.

SOURCE Outerbounds