Index Ventures leads $40 million Series B for Stackhouse AI to rein in enterprise generative AI spend

Stackhouse AI, a New York based startup that manages routing, cost controls, and compliance for enterprise generative AI workloads, raised a $40 million Series B led by Index Ventures, the company said Tuesday. Snowflake Ventures joined the round, along with existing backers Costanoa Capital and Flybridge Capital, to expand Stackhouse’s policy engine and on premises support for regulated customers.

The timing adds to a crowded Tuesday funding slate and signals that investors remain active in AI infrastructure that shows clear unit economics. The deal lands as enterprises shift from pilots to production use of large models and face scrutiny from finance and risk teams over per token costs and auditability.

Founded in 2024 by CEO Maya Chen, formerly on AWS’s pricing and cost management team, and CTO Luis Ortega, previously at Datadog, Stackhouse offers a control plane that selects among foundation models based on latency, accuracy, jurisdiction, and budget policies. The platform provides evaluation harnesses with golden datasets, enforcement for data residency, and unified audit trails that map prompts and outputs to downstream business metrics.

Customers connect OpenAI, Anthropic, Google, and open source models on AWS, Azure, or private clusters, then apply policies that constrain spend and route traffic to the lowest cost model that meets a task’s quality threshold. The company says it reduces inference costs by double digit percentages in production while maintaining service level objectives, and it publishes model performance and cost benchmarks to procurement teams.

Deal terms were not disclosed beyond the round size. The company said it will use the capital to expand European support as the EU AI Act begins to bite in 2026 and to add connectors for legacy systems in financial services and healthcare. Stackhouse sells a usage based platform fee with enterprise contracts that include compliance features and premium support.

Investor appetite for this category reflects two forces, budgets and risk. CFOs want predictability on model spend and model risk officers want traceability that satisfies existing frameworks in banking and healthcare. That favors vendors with deep integrations, clear buyer alignment, and retention tied to production usage rather than experimentation.

For founders, the bar has shifted from model novelty to operational control and compliance readiness. As model providers cut prices and add features, the defensible layer is moving to orchestration, governance, and data lineage. If this cohort executes, the near term winners in enterprise AI will be the middleware companies that turn stochastic systems into accountable software.