DevOps / SRE (ARIP) - CP Axtra

Own ARIP's operational substrate — infrastructure-as-code (on Data Engineering & Advanced Analytic (DEAA's) Terraform standard), CI/CD pipelines with eval-gate

enforcement, per-agent observability, per-agent cost meter, FinOps, and incident response. Inherit production environment from

Databricks contractor in Q1 2027 and harden it for Wave 3 scale across 15 agents and 5 suites.

Remote candidates outside of Thailand are welcome to apply.

Key Responsibilities:

• Adopt DEAA's Terraform standard for all ARIP infrastructure; author ARIP-specific modules (agent runtime, vector DB, KG database); weekly drift detection — zero unmanaged production resources
• Build ARIP CI/CD pipelines on DEAA's spine with eval-gate enforcement — no agent reaches production without eval-pass; ≤1hr deployment lead time, ≤15min rollback time
• Implement per-agent cost meter end-to-end (LLM tokens, vector DB queries, model inference) and surface to DEAA's GenAI Cost Dashboard (DTB-51)
• Stand up ARIP on-call rotation; author runbooks for every production agent and service; lead incident response; MTTR < 60min for P0/P1
• Implement ARIP cost tagging policy (team / domain / environment / agent / suite / persona) aligned with DEAA's standard; report monthly to ARIP Cost Review
• Execute Databricks contractor hand-over in Q1 2027: inherit IaC, runbooks, observability; refactor to Lotus's standards.

Requirements

• 5+ years SRE / DevOps with production ownership of AI / data-intensive or agent-based platforms
• Terraform at enterprise scale: modules, state management, drift detection, environment promotion — expert; Terraform Associate / Professional preferred
• CI/CD for ML/AI services: GitLab CI/CD or comparable with eval-gate integration; cloud (Azure preferred); AZ-500 helpful
• OpenTelemetry + Langfuse (or equivalent) for LLM observability in production; FinOps: tagging policies, per-invocation cost attribution for LLM systems
• Incident response: on-call, post-mortems, runbook authorship at senior level; rollback orchestration with quarterly game days
• Calibre: Senior DevOps / SRE from Agoda, Grab, Shopee, LINE MAN Wongnai, SCBX, KBank, or AI-native infra teams.