DevOps Engineer Intern

Company: Valura.Ai Team: Platform & Infrastructure Location: Bangalore / GIFT City / Dubai / Remote Experience: 3 to 7 years Reports to: [Head of Engineering / CTO] Employment type: Full-time

About Valura

Valura.Ai is a regulated global-investing platform. We let investors in India (IFSCA / GIFT City, LRS route) and the UAE buy US equities, bonds, funds and gold through a single AI-native product. Behind the app sits a real financial system: a double-entry ledger, multiple broker and custodian integrations, KYC/AML pipelines, an AI multi-agent research service, and a partner API used by distributors.

We run two regulated books in two jurisdictions with different rules, different brokers and different reporting obligations. Infrastructure is not a support function here; it is the thing that keeps client money reconciled and auditors satisfied.

The role

You will own how our code gets from a pull request to production, and how we know production is healthy. That means a hybrid estate: managed AWS container workloads for production, a self-hosted platform for lower environments, and a mix of Node/TypeScript and Python services. Your job is to make that estate boring, reproducible and observable.

This is a hands-on role with real ownership. You will not be filing tickets for someone else to apply.

What you will do

CI/CD and release engineering

1. Own GitHub Actions pipelines across a Turborepo / pnpm monorepo of around twenty applications plus Python services; keep build times and cache hit rates sane.
2. Maintain and harden the branch-to-environment topology across development, staging and production, with the right gates, approvals and rollbacks.
3. Reduce release risk: automated migration checks, blue/green or rolling deploys, one-command rollback.

Cloud and infrastructure as code

1. Manage AWS: container orchestration, load balancing, managed Postgres, container registry, object storage, DNS, IAM and secrets management, across more than one region.
2. Write and review Terraform; bring remaining manually provisioned resources into code with reviewable plans.
3. Operate our self-hosted environments end to end, including reverse-proxy routing, TLS and wildcard certificates, and container lifecycle.

Databases and stateful services

1. Operate Postgres, Redis and MongoDB: backups, tested restores, point-in-time recovery, connection pooling, upgrades.
2. Guard the migration path. Schema migrations that run as part of a deploy must never be able to take an environment down.

Observability and on-call

1. Own metrics, logs, traces and alerting. We have working uptime monitoring and chat-based alerting today; you will decide what the mature version looks like and build it.
2. Define SLOs for the services that matter (order placement, deposits, ledger sync, reconciliation) and alert on symptoms, not noise.
3. Build the incident process: paging, runbooks, blameless postmortems.

Security and compliance

1. Own secrets management end to end: a single managed store, no secrets in source or CI, automated rotation, clear ownership.
2. Enforce least-privilege IAM and network segmentation, and keep the public attack surface as small as the product allows.
3. Support audit and regulatory requirements in both jurisdictions: access logs, change history, evidence of controls, data residency.

Developer experience

1. Make local development fast and safe, with strong guardrails between developer environments and production data.
2. Provide ephemeral or per-branch environments where they pay for themselves.
3. Document the estate so onboarding takes days, not weeks.

What we are looking for

Must have

1. 3+ years running production infrastructure for a real product, not just lab work.
2. Strong Linux, networking and Docker fundamentals. You can debug a container that will not start, a TLS chain that will not validate, and a connection pool that is exhausted.
3. Solid AWS experience, especially container workloads (ECS or EKS), load balancing, IAM and managed Postgres.
4. Terraform in anger: modules, state management, drift, imports.
5. CI/CD ownership on GitHub Actions (or equivalent) for a multi-service repo.
6. Comfortable with Postgres operations: backups, restores, replication, migration safety.
7. Scripting in Python and/or Bash; able to read Node/TypeScript and Python application code well enough to diagnose a problem and open the fix yourself.
8. Clear written communication. Runbooks and postmortems are part of the job.

Nice to have

1. Fintech, brokerage, payments or another regulated environment; familiarity with audit and change-control expectations.
2. Self-hosted PaaS experience (Coolify, Dokku, CapRover) or bare-metal Docker fleets.
3. Kubernetes, if we decide we need it.
4. Observability stacks: Prometheus/Grafana, OpenTelemetry, Loki, Sentry, Datadog.
5. An OIDC provider such as Keycloak, plus VPN and zero-trust access tooling.
6. Cost engineering. Knowing what a workload should cost and why it does not.
7. Experience supporting Python ML/LLM services in production.

How we will evaluate you

1. Intro call (30 min) with engineering.
2. Technical conversation (60 min): walk through an incident you owned end to end, and a system you built or rebuilt.
3. Practical exercise (take-home, timeboxed to ~3 hours, or a live pairing session): containerize and deploy a small service with Terraform and a CI pipeline, then tell us how you would monitor it.
4. Systems and judgement round (60 min): architecture, migration safety, secrets management, and regulated data across regions.
5. Founder conversation.

Why this role is interesting

1. Real ownership of a production estate from day one, with the mandate to redesign the parts that need it.
2. Genuinely hard problems: two regulated jurisdictions, a double-entry ledger that must reconcile daily against external brokers, and a rapidly growing AI service tier.
3. Small team, short decision path. You will ship in your first week.
4. [Compensation band, ESOP, benefits to be inserted.]

To apply

Send your CV plus a short note on one piece of infrastructure you are proud of and one you would rebuild, to [careers@valura.ai].