07/08/2026 updated
100 % available
Senior DevOps / SRE / DBA & AI Infrastructure Engineer
Asuncion, Paraguay
Only remote
Red Hat CertifiedAbout me
Senior DevOps, SRE, AI Infra & DBA Expert. • Focus: K8s, LLMOps, High-load DBs • Impact: Up to 50% GPU cost reduction • Availability: Fractional/Part-time (≤20h/wk) • Timezone: GMT-3 (US overlap) • Payment: USDT/Crypto friendly and bank transfers
Big DataLinuxFreeBSDPython (Programming Language)PostgreSQLMySQLAnsibleApache SparkKubernetesBare MetalApache KafkaTerraformDocker
AI Infrastructure & LLMOps
Expertise in deploying and optimizing AI infrastructure including vLLM, Triton Inference Server, NVIDIA DGX clusters, NCCL, GPUDirect RDMA, InfiniBand/RoCE, NVIDIA MIG, Kubeflow, MLflow, and model quantization (AWQ/GPTQ). Proven track record of reducing Time-to-First-Token by 45% and boosting GPU utilization from 30% to 82%.
Kubernetes & Infrastructure as Code
Deep specialization in the Kubernetes ecosystem including Vanilla K8s, EKS, and GKE. Proficiency in infrastructure provisioning with Terraform and Ansible for bare-metal and private cloud environments, enabling production-ready, bulletproof cluster setups.
Database Administration (DBA)
Enterprise-grade database administration covering PostgreSQL (Patroni), MySQL, Redis Cluster, ClickHouse, MongoDB, Apache Kafka, Apache Hadoop, and Apache Spark. Extensive experience maintaining, optimizing, and scaling high-load transactional and analytical databases.
AI Agents & RAG Architecture
Design and implementation of advanced RAG pipelines with hybrid search, embedding and reranking optimization, Vector DBs (Milvus, Qdrant), Model Context Protocol (MCP), and n8n workflow automation for enterprise AI agent orchestration.
CI/CD & GitOps
Setup and management of CI/CD pipelines using GitLab CI/CD, GitHub Actions, ArgoCD, and Jenkins. Experience cutting core microservice deployment times significantly through GitOps transformation.
Telemetry & Monitoring
Building comprehensive monitoring and observability stacks using VictoriaMetrics, Prometheus, Grafana, NVIDIA DCGM Exporter, and Nagios. Skilled in tying hardware metrics to LLM inference metrics and reducing alert fatigue.
OS & Networking
Expert-level Linux administration, FreeBSD, OpenStack, and low-latency performance tuning. Experience optimizing Linux kernel configurations and network stacks for high-load, low-latency environments.
Cloud & Multi-Cloud Infrastructure
Experience developing multi-cloud infrastructure on AWS EKS and GCP GKE using Terraform, transitioning monolithic applications into containerized environments and managing private cloud infrastructure with OpenStack.
Big Data Infrastructure
Administration of Big Data infrastructure architectures using Apache Hadoop and Apache Spark, including deployment and management for large-scale data processing workloads.
Expertise in deploying and optimizing AI infrastructure including vLLM, Triton Inference Server, NVIDIA DGX clusters, NCCL, GPUDirect RDMA, InfiniBand/RoCE, NVIDIA MIG, Kubeflow, MLflow, and model quantization (AWQ/GPTQ). Proven track record of reducing Time-to-First-Token by 45% and boosting GPU utilization from 30% to 82%.
Kubernetes & Infrastructure as Code
Deep specialization in the Kubernetes ecosystem including Vanilla K8s, EKS, and GKE. Proficiency in infrastructure provisioning with Terraform and Ansible for bare-metal and private cloud environments, enabling production-ready, bulletproof cluster setups.
Database Administration (DBA)
Enterprise-grade database administration covering PostgreSQL (Patroni), MySQL, Redis Cluster, ClickHouse, MongoDB, Apache Kafka, Apache Hadoop, and Apache Spark. Extensive experience maintaining, optimizing, and scaling high-load transactional and analytical databases.
AI Agents & RAG Architecture
Design and implementation of advanced RAG pipelines with hybrid search, embedding and reranking optimization, Vector DBs (Milvus, Qdrant), Model Context Protocol (MCP), and n8n workflow automation for enterprise AI agent orchestration.
CI/CD & GitOps
Setup and management of CI/CD pipelines using GitLab CI/CD, GitHub Actions, ArgoCD, and Jenkins. Experience cutting core microservice deployment times significantly through GitOps transformation.
Telemetry & Monitoring
Building comprehensive monitoring and observability stacks using VictoriaMetrics, Prometheus, Grafana, NVIDIA DCGM Exporter, and Nagios. Skilled in tying hardware metrics to LLM inference metrics and reducing alert fatigue.
OS & Networking
Expert-level Linux administration, FreeBSD, OpenStack, and low-latency performance tuning. Experience optimizing Linux kernel configurations and network stacks for high-load, low-latency environments.
Cloud & Multi-Cloud Infrastructure
Experience developing multi-cloud infrastructure on AWS EKS and GCP GKE using Terraform, transitioning monolithic applications into containerized environments and managing private cloud infrastructure with OpenStack.
Big Data Infrastructure
Administration of Big Data infrastructure architectures using Apache Hadoop and Apache Spark, including deployment and management for large-scale data processing workloads.
Experience
Professional experience10+ years
Freelancer since2021
Languages
EnglishGoodRussianNative speakerSpanishBasic knowledge
Project history
Architected and deployed an on-premise Enterprise AI platform within a sovereign private cloud for customer support chatbots, voice assistants, and scoring analytics. Designed and deployed a bare-metal NVIDIA DGX computing cluster with NCCL and GPUDirect RDMA over 100G/200G Mellanox InfiniBand/RoCE. Deployed vLLM and Triton Inference Server stacks in Kubernetes, reducing TTFT by 45%. Configured NVIDIA MIG in K8s boosting GPU utilization from 30% to 82%. Engineered fault-tolerant enterprise RAG systems with hybrid search using Milvus Vector DB and HashiCorp Vault. Built CI/CD pipelines for automated model quantization via GitLab CI and ArgoCD.
Maintained on-premise Vanilla K8s bare-metal clusters for a high-load betting platform ensuring 99.99% SLA uptime. Revamped GitLab CI/CD pipelines and introduced ArgoCD, cutting core microservice deployment times in half. Deployed vLLM and Triton Inference Server in K8s for anti-fraud scoring and user support. Configured NVIDIA MIG on bare metal, increasing video card utilization by 40%. Integrated Qdrant Vector DB for customer care RAG systems and replaced legacy automation scripts with n8n workflows in K8s.
Migrated private cloud infrastructure (OpenStack) entirely to Terraform, reducing environment setup times from days to 15 minutes. Maintained platform stability during major sports traffic spikes (5x-10x normal load) with Linux kernel and network stack optimizations. Designed and automated fault-tolerant PostgreSQL clusters using Patroni and distributed Redis Clusters. Built a Prometheus and Grafana monitoring suite with custom Kafka queue metrics, reducing alert fatigue by 40%. Implemented self-hosted n8n for infrastructure self-healing.