AIOps Engineer

DevOps/AIOps Engineer (Platform)

Experience: 3–5 Years

About the Company

We aim to bring about a new paradigm in medical image diagnostics — intelligent, holistic, ethical, explainable, and patient‑centric. We're looking for innovative problem‑solvers who empathize with clinicians and patients, understand business problems, and can design and deliver reliable, intelligent products.

Key Responsibilities

• CI/CD for services & models: Own pipelines (GitHub Actions/GitLab CI), environment gates, artifact/version governance (containers, models, SBOMs), safe rollouts & instant rollbacks.
• Kubernetes platform (EKS preferred): Operate multi-env clusters; Helm/Kustomize; GitOps (Argo CD/Flux); progressive delivery (canary/blue-green/Argo Rollouts/Flagger).
• Serving & APIs: Deploy and tune FastAPI services and Triton/ONNX/TensorRT inference; traffic shaping, runtime config, autoscaling signals.
• Event-driven orchestration: Build robust consumers/producers on RabbitMQ/ActiveMQ/Kafka with back-pressure, dead-lettering, idempotency, and retry patterns.
• Observability & AIOps: Define SLIs/SLOs and error budgets; metrics/logs/traces (Prometheus/Grafana/Loki/Tempo/ELK); intelligent alerting & noise reduction; basic model/data drift hooks.
• Security in SDLC: Supply-chain security (image signing/provenance, SBOM scans), SAST/DAST/IaC scanning, policy-as-code (OPA/Gatekeeper), secrets hygiene in pipelines/workloads.
• Data/Model platform integration: S3/MinIO for artifacts; integrate model registry (MLflow or similar) into CD; immutable, traceable releases.
• Resilience & performance: Capacity planning (incl. GPU), autoscaling (HPA/VPA/KEDA), caching/queue tuning; chaos/game-days; write runbooks and own incident response for platform services.
• Developer experience: Golden paths, starter repos, internal Helm charts, docs & enablement to make shipping boring and fast.
• FinOps mindset: Cost dashboards, right-sizing, bin-packing, GPU utilization policies, spot vs on-demand strategy.

Skills and Qualifications (Required)

• 3+ years in DevOps/SRE/MLOps with strong Docker & Kubernetes fundamentals.
• Production CI/CD expertise; canary/blue-green; artifact & version management.
• IaC (Terraform) and GitOps workflows (Argo CD/Flux).
• Observability: Prometheus/Grafana; logs/traces with Loki/Tempo/ELK.
• Production message queues (RabbitMQ/ActiveMQ/Kafka) with back-pressure & retries.
• Cloud experience (AWS/GCP/Azure), EKS preferred; object storage (S3/MinIO); model registries (MLflow or similar).
• Security in SDLC and compliance guardrails for PHI-like data (least-privilege IAM, secrets, auditability).
• Incident response experience; writing SLIs/SLOs, runbooks, and operating to error budgets.
• Scripting for platform tasks (Python/Bash).

Preferred

• Triton Inference Server, ONNX/TensorRT optimizations; GPU scheduling on K8s (NVIDIA device plugin, MIG, node pools).
• Argo Rollouts/Flagger, Karpenter, KEDA; caching layers (Redis/NVCache patterns).
• Policy-as-code (OPA/Gatekeeper), image signing (cosign), SBOM tools (syft/grype).
• Network savvy for app delivery (ingress, service meshes, egress policies).

Education

BE/B.Tech or equivalent experience.

Location & Work Setup

On-site - Gurugram

Job Type: Full-time

Pay: ₹800, ₹1,000,000.00 per year

Application Question(s):

• How many years of hands on production kubernetes experience do you have?
• Which CI/CD tools have you used to deploy micro services?
• Do you have experience with progressive delivery and roll backs?
• What is your experience with serving ML Models in production?
• How do you evaluate and track the performance of a Model in production?

Work Location: In person