CYNET

Platform Engineer

AI-Native Infrastructure • GCP • Healthcare Staffing Technology

Location: Remote / Hybrid Team: AI Engineering (~25) Type: Full-time (open to fractional/contract-to-hire)

We are an AI-native engineering org building internal tooling and AI products for healthcare and technology staffing. You will own our cloud platform on Google Cloud and build the paved road” that lets our developers —

and our AI coding agents — ship to production safely and fast. This is the senior, leveraged evolution of a DevOps role: less pipeline babysitting, more building self-service infrastructure and guardrails. Because we handle protected health information (PHI), HIPAA-grade access control and auditability are core to the job, not an afterthought.

WHAT YOU’LL OWN

- Cloud platform — Design, provision, and operate our GCP estate (Cloud Run, GKE/Autopilot, BigQuery,

Artifact Registry, networking, secrets) as code.
- Paved road — Build an internal developer platform: golden CI/CD paths, preview environments, and

one-command deploys so engineers self-serve without filing tickets.
- Agentic infra — Stand up and harden the infrastructure our agents run on — sandboxes, MCP server

deployments, and the CI/CD surfaces agents use to open and merge PRs.
- Security & compliance — Enforce least-privilege IAM, audit logging, secrets hygiene, and HIPAA / SOC 2

controls across all environments.
- Reliability & cost — Own observability and cost: cloud spend (FinOps) and LLM token spend, with

dashboards, alerting, and SLOs. MUST-HAVE QUALIFICATIONS
- 4+ years in DevOps / SRE / Platform Engineering,



with production ownership (not just CI tweaks).
- Deep GCP experience (Cloud Run, IAM, BigQuery, VPC, Secret Manager). Equivalent AWS/Azure depth +

willingness to go all-in on GCP is acceptable.
- Robust Infrastructure-as-Code: Terraform (required), plus containers (Docker) and CI/CD via GitHub Actions.
- Has built self-service tooling or internal platforms that other engineers actually used.
- Comfortable working in an AI-heavy workflow — uses coding agents daily and understands how they touch

infra. NICE-TO-HAVE
- Kubernetes depth (CKA), experience with MCP servers / agent sandboxes (E2B, Daytona, Cloud Run jobs),

and LLM observability (Helicone, Langfuse).
- Prior work in a regulated / PHI or PII environment (healthcare, fintech).

RECRUITER SCREENING CHECKLIST Use this to qualify candidates before they reach the hiring manager. A strong candidate clears every Hard Filter and most Signals.

Hard filters (disqualify if missing)

Terraform in production. Ask: “Walk me through infra you manage as code today.” Vague answers = no.

GCP (or strong cloud) depth. Must name real services they’ve run: Cloud Run, IAM, BigQuery,

VPC, Secret Manager.

CI/CD ownership.



Has built pipelines in GitHub Actions (or GitLab/Cloud Build), not just consumed them.

Containers. Docker fluency; can explain an image build and a deployment they shipped.

Regulated-data awareness. Can speak to least-privilege IAM, secrets handling, and audit logging. HIPAA/SOC 2 a plus.

Strong signals (the differentiators)

Built an internal platform. Made deploys/infra self-service for other engineers — ask for a concrete example.

Works with AI coding agents. Uses Claude Code / Cursor / Codex daily; understands sandboxes and agent-driven PRs.

MCP / agent infrastructure. Has deployed or operated MCP servers or sandbox environments

(E2B, Daytona, Cloud Run jobs).

Cost ownership. Has driven down cloud spend (FinOps) and/or tracked LLM token cost.

Kubernetes. CKA or real GKE production experience (only if we run GKE).

Three phone-screen questions

- “Tell me about the last internal tool or platform you built so other engineers could self-serve infrastructure.

What did they used to do manually?”
- “How do AI coding agents fit into your workflow today, and how does that change what infrastructure needs to

provide?”
- “You’re given a GCP project handling patient data. What are the first five things you lock down?” (Listen for

IAM, secrets, audit logs, VPC, encryption.)

Red flags

- Only consumed CI/CD others built; never owned infra.
- Resume is all certifications, no production stories.
- Dismissive of AI agents in the dev workflow.
- Can’t articulate a least-privilege / audit approach to sensitive data.

📌 Devops Engineer (Dehradun)
🏢 Cynet Systems
📍 Dehradun