Ellamind is building an enterprise-grade platform for AI evaluation and optimization, with a tech stack anchored in Python (Django, FastAPI), PyTorch, and inference optimizers (vLLM, SGLang, Megatron-LM). The founding team and active projects reveal a company in transition: evaluation infrastructure and experiment management dominate the roadmap, while pain points around structured outbound sales and multi-tenant scaling suggest they're shifting from product-first to go-to-market—backed by engineering and research talent but with sales infrastructure still being built.
Notable leadership hires: Tech Lead, Sales Lead
Ellamind, founded in 2024 and headquartered in Bremen, Germany, builds a platform to simplify AI evaluation and optimization for enterprise customers. The product centers on evaluation infrastructure, experiment management, and a lightweight inference engine, with architecture designed for distributed systems and both cloud and on-premise deployment. The team of 11–50 is heavily weighted toward engineering (with distributed systems and ML optimization expertise), supported by early sales and research functions. Active hiring is concentrated in Germany, focusing on senior and lead-level engineering roles.
Ellamind's stack spans Python (Django, FastAPI), JavaScript/TypeScript (Next.js, React), PostgreSQL, Kubernetes, and ML frameworks including PyTorch, Hugging Face, vLLM, and SGLang for model inference optimization. Docker, Celery, Ray, and SLURM handle distributed compute.
Core projects include evaluation infrastructure, experiment management, a lightweight commercial inference engine, distributed system architecture, and multi-tenant deployment capabilities. Concurrent efforts cover automated testing frameworks, AI model integration, and sales process design.
Other companies in the same industry, closest in size