CoreWeave Launched Forge Platform for Agentic AI
The new platform integrates disparate AI workflows to accelerate model development and iteration cycles.
Updated on Oct. 9, 2026 in Artificial Intelligence
Live Poll
Do you believe it is feasible to continuously improve AI agents after initial production deployment?
CoreWeave has announced the general availability of its Forge platform, designed to unify business and machine learning operations for agentic AI. The platform features an AI Research and Iteration Agent (ARIA) that assists developers by identifying patterns for model distillation.
Why it matters
The platform aims to reduce the time required for AI model learning and deployment by streamlining the technical workflow. It consolidates disparate development tools into a centralized pipeline for teams building complex autonomous agents.
Forge utilizes a five-stage loop—run, observe, curate, improve, and evaluate—to automate model training. Cognition AI Inc. is already running production workloads on the system using CoreWeave's Nvidia Vera Rubin NVL72 hardware racks.
The players
CoreWeave
A specialized cloud infrastructure provider focused on high-performance compute and large-scale GPU clusters for AI workloads.
Cognition AI Inc.
An AI development lab focused on building autonomous software engineers capable of performing complex coding and reasoning tasks.
The details
Forge connects fragmented tools into a single platform that handles agent development. The system uses ARIA to perform model distillation—a process where a smaller model learns to mimic the performance of a larger, more complex model—by identifying patterns within training data. By standardizing these stages, the platform allows machine learning teams to move through the improvement loop more efficiently.
Timeline
October 8, 2026: CoreWeave announced the Forge platform and the general availability of ARIA.
The Tech Race
Forge integrates with hardware like the Nvidia Vera Rubin NVL72 architecture to optimize the underlying infrastructure for agentic AI workloads. This strategy marks a shift toward vertical integration where cloud providers align their software development environments closely with the high-performance hardware required for model distillation.
Business units and machine learning teams can now utilize a shared workflow to reduce friction when developing and deploying AI agents. The platform is currently available to users seeking to build on CoreWeave's hardware stack.
The takeaway
CoreWeave is shifting the focus of cloud providers from raw compute to integrated workflows that automate the AI lifecycle. Readers should monitor future partner network announcements to see if these co-engineered integrations expand beyond current early adopters like Cognition AI Inc.
Further reading
For more on the latest trends in autonomous infrastructure, see our coverage in Artificial Intelligence.
Live Poll
Do you believe it is feasible to continuously improve AI agents after initial production deployment?
