Keep pulling the thread on CJ Desai.
A top 50 customer of a major hyperscaler, located in Texas, was denied additional public cloud capacity for new workloads, including AI, due to the provider's capacity constraints.
A large Fortune 100 customer in Texas reversed its decision to decommission a data center and will now run workloads, including AI, on-premises due to a lack of available public cloud capacity.
ElevenLabs runs over 50 million agents on MongoDB.
The demand for on-premise data centers is being driven by two main factors: hyperscalers running out of capacity and enterprises' desire to keep proprietary data on their own premises.
Enterprises are not standardizing on a single type of LLM; instead, they use a mix of open-source, closed-source, small, large, and domain-specific models from providers like OpenAI and Anthropic depending on the specific use case.
A frontier AI lab chose MongoDB as its database because of MongoDB's ability to operate across multiple clouds, which the lab valued for resiliency.
A frontier AI lab uses MongoDB for inference, voice products, and video and image creation workloads.
Major hyperscalers have all announced "forward deployed engineer" programs in the last 8 weeks, a strategy similar to Palantir's playbook.
The ability for software to write software has significantly lowered both the incremental cost of innovation and the cost of experimentation.
A large U.S. telecommunications company was forced to adopt a multi-cloud strategy after its primary hyperscaler partner refused to provide needed additional regional capacity.
Multiple frontier AI labs use MongoDB for a variety of use cases.
AI-native startups including Emergent, Base44, Metal.AI, and ElevenLabs are built on MongoDB.