AI infrastructure and LLM operations for lifestyle e-commerce at scale
Prosus operates a large-scale AI and ML infrastructure layer serving lifestyle e-commerce across Europe, Latin America, and India. The tech stack signals heavy investment in LLM serving (vLLM, Megatron-LM, DeepSpeed) and orchestration (Kubernetes, Argo CD, Kubeflow), with active work on domain-specific model training, RLHF pipelines, and agentic systems. Engineering hiring is accelerating (6 roles posted in 30 days) skewed toward senior-level, paired with parallel scaling in security and legal—suggesting Prosus is managing AI deployment risk alongside rapid infrastructure build-out.
Prosus is a public e-commerce technology company headquartered in Amsterdam building infrastructure and AI ecosystems for lifestyle brands across Europe, Latin America, and India. The platform unifies commerce, fintech, food, and experiences layers, serving both buyer and seller sides. The operational footprint spans infrastructure engineering (LLM deployment, distributed training, A/B testing and canary systems), compliance and governance programs, and emerging regulatory tooling. Current focus areas include GPU cost optimization, latency-throughput trade-offs in inference workloads, and bridging gaps between rapid AI innovation cycles and end-user adoption.
Prosus runs vLLM, Kubernetes, Slurm, Python, FastAPI, MLflow, Weights & Biases, Ray, DeepSpeed, Megatron-LM, GPT-4, Claude, and AWS/Azure/GCP across infrastructure and ML operations. GitHub and Argo CD handle CI/CD.
Active projects include LLM serving infrastructure, domain-specific language model training, RLHF post-training pipelines, distributed training infrastructure, and an agentic work platform. Parallel work spans A/B testing, model evaluation, and compliance tooling.
Other companies in the same industry, closest in size