Platform & Infrastructure Engineer

We are looking for a Platform & Infrastructure Engineer based in Latin America to work on a long-term project for one of our clients, a Software Development company based in Atlanta, Georgia.

Our client is building a secure, bank-grade AI infrastructure and developing cutting-edge applications designed to power the next generation of modern banking and financial services.

• Own the cloud foundation across Azure and GCP using infrastructure as code, covering networking, identity and access, Kubernetes clusters, storage, and secrets, while meeting the stringent security standards expected by banking customers.
• Build and evolve the Kubernetes platform: GPU node pools for model serving, autoscaling, upgrades, ingress, and the stateful and AI workloads that run on it.
• Run the data layer on the platform, from the graph database behind the banking ontology to relational databases and job pipelines, with tested backup, restore, and disaster recovery procedures.
• Build the environment blueprint: new tenants and dedicated environments created from one codebase in hours, not weeks.
• Make our client’s platform installable in a bank's own cloud: a versioned, signed release with a clear upgrade path and runbooks a bank's engineers can follow, including restricted networks.
• Build the observability foundation, the shared stack every team builds on, with metrics, logs, traces, dashboards, and alerts defined as code, so every environment is visible the same way.
• Own compute and GPU capacity and cost, choose the tools across infrastructure and data, troubleshoot production issues, and take part in on-call.

• Advanced Level of English.
• 5+ years in infrastructure or platform engineering, running production systems for other engineers and customers.
• Expertise in infrastructure as code:
  Terraform in a team setting, with modules, remote state, code review, and production applies across many environments.
• Hands-on experience operating Kubernetes clusters, including upgrades, networking, scheduling, storage, and troubleshooting pod startup issues.
• 2+ years of experience running GPU workloads and model serving in production on Kubernetes, including GPU node pools and drivers, GPU sharing and scheduling, autoscaling, and serving frameworks such as vLLM.
• Experience operating a graph database in production (Neo4j, ArcadeDB, Dgraph, Janus Graph, or similar), along with relational databases: sizing, backup and restore, upgrades, and performance under load.
• Deep knowledge of at least one major cloud, Azure, AWS, or GCP, including private networking, DNS, identity, and security, with least privilege access and change control practices that hold up to a customer review.
• Experience deploying and operating software in environments with limited control, including dedicated tenants, customer clouds, on-premises infrastructure, or air-gapped environments.
• Strong programming skills in Python, Go, or similar languages, with experience building tools and automation rather than relying solely on configuration.
• Strong troubleshooting skills, with the ability to trace issues through logs, metrics, traces, and Kubernetes state to identify the root cause.
• Strong willingness to learn and build, sound judgment on technical trade-offs, clear written communication, and experience using AI coding tools while taking ownership of the resulting code.

• Bachelor’s Degree in Computer Science, Systems Engineering or related fields.