DevOps / Site Reliability Engineer
Kaspi Bank
Kazakhstan-based bank and fintech company.
Oct 2025 — PresentOperates production Kubernetes and GPU platforms for AI/ML and fintech workloads, including infrastructure supporting 99.9% SLA and a Milvus database with 1M+ biometric embeddings.
• Operate 10+ production Kubernetes clusters (20–40 nodes each) supporting biometric authentication, computer vision, LLM, fraud detection and payment platforms. • Manage enterprise GPU infrastructure with 50+ NVIDIA accelerators (B200, H200, L40S), including NVIDIA MIG partitioning, workload placement and GPU resource allocation. • Deploy, operate and troubleshoot 100+ production AI/ML workloads, including FastAPI services, Triton Inference Server, LLM inference platforms and computer vision services.
