LangSmith BYOC lands on AWS, Azure and Google Cloud: data stays in your account, updates stay with LangChain
Synopsis
LangChain has extended LangSmith's BYOC deployment from AWS to Azure and Google Cloud, with AWS generally available and Azure and Google Cloud in Beta. Until now, teams with strict compliance boundaries ran self-hosted LangSmith, which required dedicated platform engineers to patch stateful services, configure databases and run schema migrations; BYOC keeps the control plane with LangChain and places the data plane holding traces, prompts, datasets and agent deployments in the customer's own cloud account and virtual network, while LangChain handles scaling, schema updates and runtime dependencies.
Interpretation
LangSmith's BYOC deployment now covers AWS, Azure and Google Cloud, letting customers keep traces, evals, datasets, prompts and agent deployments in a cloud account of their choosing while LangChain operates the data plane. Before this, teams with strict compliance boundaries deployed and maintained self-hosted LangSmith; data stayed inside their boundary, but it took dedicated platform engineers to patch stateful services, configure databases and run schema migrations. Since BYOC launched on AWS in August, many teams have chosen to run LangSmith in their own cloud with BYOC, citing three benefits: continuous upgrades, faster access to the latest LangSmith features, and automatic scaling with load.
The split separates control plane from data plane: the LangChain-run control plane holds auth, org and workspace config, billing metadata, hashed API keys and static UI assets, while the customer-owned data plane holds sensitive content such as traces, prompts, datasets, experiments, agent deployments and workspace secrets. An organization can hold several data planes, each in a specific account and region and each holding its own workspaces, so data planes can be split by region, environment or business unit. For example, a platform team might run production and staging data planes in separate AWS accounts, while a business unit standardized on Azure runs its own separate BYOC deployment on Azure, with each team's traces landing in the account that team owns.
Perspective
The demonstrated scope is LangSmith BYOC generally available on AWS and in Beta on Azure and Google Cloud, with the data plane running inside the customer's own virtual network and account while LangChain handles scaling, schema updates and runtime dependencies.
BYOC on Azure and Google Cloud is currently in Beta; Google Cloud data planes can run in any region at launch but are managed from LangSmith's US control plane; LangSmith Sandboxes support on Azure is coming soon; trace migration is available on AWS today.