Krea runs a compute-intensive generative AI platform serving 30M+ users with image, video, and 3D generation tools. The tech stack reveals a mature ML operations foundation: Kubernetes + AWS/GCP/Azure for distributed training, PyTorch + DuckDB for model work, and observability via Prometheus + OpenTelemetry. Active projects center on foundation model training, large-scale GPU orchestration, and real-time inference optimization — pain points around multi-thousand-node Kubernetes scaling and petabyte data pipelines signal they're operating at significant infrastructure complexity.
Notable leadership hires: Social Media Lead
Krea is a generative AI creative platform and research lab based in San Francisco serving designers, marketers, agencies, and product teams. The product enables rapid generation and editing of images, video, and 3D assets, with emphasis on creative control and brand consistency. The company operates at scale: 30M+ users and a tech stack spanning Kubernetes orchestration, multi-cloud deployment (AWS, GCP, Azure), and GPU-heavy ML workloads. The team is engineering-forward with lean go-to-market — 5 engineers, 1 salesperson, and active hiring focused on mid- and senior-level technical roles.
Krea's core stack includes Kubernetes and multi-cloud (AWS, GCP, Azure), Python and Rust for backend, JavaScript/TypeScript/Svelte for frontend, PostgreSQL + Redis for data, PyTorch for ML training, and observability via Prometheus and OpenTelemetry.
Active projects include foundation diffusion model training, large-scale Kubernetes cluster management for ML, fault-tolerant inference systems, real-time image generation optimization, and video processing at scale across millions of scenes.
Other companies in the same industry, closest in size