AI Platform SRE for Regulated Environments

Archer; Boca Raton (FL); On-site; USD 120,000 - 180,000; Full time

Job summary

Archer, in collaboration with Leo Technologies, seeks a DevOps / SRE to deploy, operate, and harden AI systems used for corrections operations and intelligence analysis. You will own the path from model prototype to production, ensuring uptime, data segregation, and auditability in CJIS/FedRAMP-aligned environments.

The role requires designing LLM deployment platforms, building robust CI/CD for ML, and managing GPU-based inference infrastructure with strong security and observability practices.

Qualifications

• 5+ years in DevOps, SRE, or platform engineering.
• 2+ years operating ML/AI workloads in production.

Responsibilities

• Design and operate deployment platform for LLM apps, agentic systems, and RAG pipelines across clouds, on-premises and air-gapped environments.
• Build CI/CD pipelines for model and application delivery, including registries, versioning, evaluation gates, and rollback paths.
• Stand up and maintain inference infra: GPU clusters, model serving, vector/graph DBs, and data services.
• Operate Kubernetes as the backbone for AI workloads with GPU scheduling and isolation.
• Implement observability for AI systems: token throughput, model drift, logs, and cost per request.
• Harden environments to CJIS, FedRAMP, StateRAMP, or DoD controls with encryption and audit logging.
• Enforce data segregation and need-to-know access via IAM and secrets management.
• Build deployment patterns for air-gapped or classified enclaves with signed artifacts.
• Manage incident response: runbooks, on-call rotations, blameless postmortems, and LLM failure modes.
• Collaborate with data scientists, security, and compliance to ship safely.

Skills

Kubernetes

CI/CD

Python

Bash

Terraform/Pulumi

Security posture

Observability

GPU infrastructure

Tools

Terraform

Pulumi

vLLM/TGI/Triton

Ollama

TensorRT-LLM

Kubernetes (EKS/AKS/GKE)

GitHub Actions

GitLab CI

Jenkins

ArgoCD

Prometheus

Grafana

OpenTelemetry

Loki

Elastic/Datadog

Job description

Archer, in collaboration with Leo Technologies, seeks a DevOps / SRE to deploy, operate, and harden AI systems used for corrections operations and intelligence analysis. You will own the path from model prototype to production, ensuring uptime, data segregation, and auditability in CJIS/FedRAMP-aligned environments.

The role requires designing LLM deployment platforms, building robust CI/CD for ML, and managing GPU-based inference infrastructure with strong security and observability practices.