Overview

Leena AI is a leader in Agentic AI for the enterprise.

Responsibilities

• Leena AI runs a multi-tenant agentic AI platform for large global enterprises across fifteen production regions on AWS, Azure and GCP, plus private-cloud deployments inside customer environments. This role owns the infrastructure all of that runs on: Kubernetes clusters, cloud accounts, networking, CI runners, the observability stack, secrets and supply-chain security, and the cloud bill.
• You lead a team of 3–5 DevOps engineers & SREs. You work alongside our Platform Engineering group, which owns service-level objectives, alerting and incident command for the product; your team is the infrastructure counterpart that keeps the ground solid and makes it cheaper and safer every quarter.
• This is a hands-on management role. You will still write Terraform and review Helm charts. You will also carry a second product: our deployment package for private-cloud and OEM partners (Terraform, Helm, signed images, SBOM, preflight and post-install verification), which turns runs in our SaaS into runs in the customer's account, and they trust it.
• Own production infrastructure across fifteen regions on AWS, Azure and GCP: Kubernetes clusters (EKS/AKS/GKE), node pools including GPU nodes for in-platform ML inference, networking, DNS, TLS, CDN/WAF, and cluster-level Helm.
• Own the data-layer infrastructure our services depend on: MongoDB, Elasticsearch, Redis, RabbitMQ, ClickHouse and object storage; capacity planning, upgrades, backups and restore drills.
• Operate the observability stack (metrics, logs, traces) as a reliable, cost-controlled service for every engineering team.
• Own CI runners and build infrastructure; keep pipelines fast, reproducible and secure.
• Standardise environments through infrastructure-as-code. Every region should be reproducible from a repo, and no change should reach production by hand.
• Own the installable Leena package for private-cloud and OEM partner deployments: Terraform modules, Helm charts, signed container images, monthly SBOM regeneration, preflight checks, post-install verification and support bundles.
• Deliver customer-environment installs (for example, into a customer's own Azure or AWS subscription) with a repeatable runbook and a predictable timeline, and run the upgrade path afterwards.
• Work with Security, Legal and the partner's engineering team on install requirements, egress allow-lists, air-gap constraints and evidence for customer audits.
• Own a seven-figure annual cloud bill. Baseline it per region and per tenant tier, then reduce it: right-sizing, commitments, storage tiering, egress, and retiring what nobody uses.
• Report unit economics monthly (cost per region, per tenant, per conversation) with the numbers stated from memory, not from a dashboard you have to open.
• Own secrets management (Vault/KMS), image signing, vulnerability remediation SLAs, CIS hardening, network segmentation and least-privilege IAM across all clouds.
• Provide infrastructure evidence for SOC 2, ISO 27001 and customer VAPT/security reviews; close infrastructure findings on time.
• Execute the security tooling roadmap (scanning, runtime protection, audit logging) in partnership with the Security function.
• Partner with Platform Engineering on reliability: infrastructure-side root cause, capacity headroom, failover and DR testing, and closing postmortem actions that land on infrastructure.
• Build a DevOps on-call tier that responds fast and pages rarely; aim for a quiet rotation, and treat recurring pages as bugs to be engineered away.
• Lead, mentor and grow the DevOps team; hire the SRE; set clear ownership so every cluster, account, pipeline and bill has a named owner.
• Replace hero work with runbooks, automation and documentation (Confluence). If a task needs one specific person, that is a risk to remove.
• Run the roadmap like an engineering team: quarterly goals, a visible backlog, and monthly reporting to the SVP Engineering on reliability, cost and security posture.

Qualifications

• Experience: 8–12 years in DevOps, SRE, cloud or infrastructure engineering, including 2+ years managing engineers. Prior experience as a strong IC who still enjoys the keyboard.
• Kubernetes at depth: Operated multi-cluster, multi-region Kubernetes in production, including upgrades, autoscaling, networking (CNI, ingress, service mesh or equivalent) and stateful workloads.
• Multi-cloud: Production ownership on at least two of AWS, Azure and GCP; IAM, networking and cost models of each.
• Infrastructure as code: Expert in Terraform and Helm; disciplined about modules, state, reviews and drift.
• CI/CD & automation: Built and operated CI/CD at scale (GitHub Actions, GitLab CI, ArgoCD or similar). Fluent in Python, Go or Bash.
• Observability operations: Run Prometheus/Grafana, ELK/OpenSearch, OpenTelemetry or comparable stacks as services with their own cost and retention decisions.
• Cost ownership: Has owned and materially reduced a cloud bill, and can walk through how, with numbers.
• Security fundamentals: Secrets management, image signing and SBOM, vulnerability management, CIS benchmarks, and audit evidence for SOC 2 or ISO 27001.
• Incident discipline: Has been a responder in production incidents, written postmortems, and driven the fixes to closure.
• Communication: Explains trade-offs (cost vs. resilience, speed vs. control) plainly to engineers, executives and customers.
• Education: Bachelor's degree in Computer Science, Engineering, or a related field.

Nice to have

• Shipped software installed into customer-controlled environments: private cloud, on-prem, air-gapped or OEM/partner packaging.
• Operated GPU infrastructure for ML or LLM inference workloads.
• Production operation of MongoDB, Elasticsearch, Redis, RabbitMQ or ClickHouse at scale.
• Multi-tenant B2B SaaS serving large enterprises, with data-residency and regional isolation requirements.
• Hands-on experience through a compliance audit cycle (SOC 2 Type II, ISO 27001) as the infrastructure owner.
• FinOps practice: commitment planning, showback/chargeback, unit cost reporting.