Description
Responsibilities:
Infrastructure Ownership: Own Helpshift production services and ensure complete monitoring coverage, troubleshoot and fix production issues.
Infrastructure as Code (IaC): Design and maintain scalable GCP infrastructure using Terraform o
AI Orchestration & LLMOps: Build deployment pipelines for AI agents, managing vector databases (e.g., Vertex AI Search, Pinecone, Weaviate, ElasticSearch) and model endpoints.
Security (DevSecOps): Implement "Security-by-Design," including IAM least-privilege access, secret management (Secret Manager), and automated vulnerability scanning for AI workloads.
CI/CD Excellence: Architect high-velocity pipelines for both traditional microservices and AI model prompts/configurations. Design, implement, and maintain secure CI/CD pipelines for automating deployment, configuration, and testing processes.
Observability: Set up comprehensive monitoring for system health and LLM-specific metrics (latency, token usage, and cost)
Cloud Governance: Optimise GCP costs and manage resource quotas,



especially for GPU/TPU-intensive AI tasks.
Cross Cloud Deployment: Establish & Optimise the connectivity among apps deployed in different cloud settings (AWS GCP)

Requirements
Requirements
Relevant experience of 6+ years and above
Expert-level Google Cloud Platform (GCP) administration skills: GKE, Cloud Run, Vertex AI, GCS, NEG etc
Experience deploying Vector Databases (Pinecone, Weaviate, ElasticSearch or Vertex Search) and managing API rate limits/throttling for LLM providers.
Setting up Cloud Monitoring/Logging specifically for AI metrics: token consumption, inference latency, and model error rates.
In-depth knowledge of running/managing UNIX-like operating systems (we use Ubuntu)
Robust knowledge of networking protocols, security architectures, and identity and access management (IAM) principles.
Experience with containerisation technologies (e.g., Docker, Kubernetes) and securing containe

📌 Devops Se Ii Pune (India)
🏢 Keywords Studios
📍 India