Backend and Infrastructure Engineer

About the job

You will join our product engineering team on a live consumer mobile product used across North America, and own its backend: Python services on AWS with AI services in the path of every user session. We build the product, run it in production, and decide how it evolves.

One person owns all of it: the services, the APIs, the vendor integrations, and the AWS account they run on. Roughly half the job is backend development and half is infrastructure, deployment, and operations. There is no separate DevOps engineer. If you want to write application code and hand the cloud to someone else, this is the wrong seat.

You work alongside our senior mobile lead, who holds architecture, and our QA engineer. You make sound calls on your own, with the tech lead there to check them against, not to make them for you. Your first real task is reading an undocumented system and reconstructing what it actually does.

This is not an applied ML role. Model training and evaluation research are a separate function. This role builds the plumbing that makes that possible: reliable integrations, honest cost and usage data, and a system someone can safely change.

This is a contract role, full-time or part-time depending on fit. We agree which one with you during the interviews, and a part-time contract can move to full-time as the work grows. It is hybrid: you work from home most of the month and from our Gurugram office 3–4 days a month. Gurugram is preferred, but we will consider strong candidates elsewhere in India who can travel for those days. Rate is benchmarked to seniority and we share a number in the first conversation.

Responsibilities

Backend and AI integrations

• Build and maintain the Python services, APIs, and Lambda functions behind the app, and debug issues across development and production.
• Integrate the AI vendors through a clean integration layer rather than vendor logic scattered through the application: authentication, validation, timeouts, capped retries, rate limits, and malformed responses.
• Design fallback and degraded behavior for when a vendor is slow or down, including when one fails and the others succeed.
• Instrument cost, latency, and usage per request and make it queryable, and keep AI spend under control.
• Write SQL against production data to answer real product questions, and handle sensitive user data carefully: retention, deletion, and what leaves the system.

Infrastructure and operations

• Own and extend the Terraform estate, including state and environment separation.
• Stand up a staging environment and a repeatable, reversible deployment pipeline.
• Establish IAM fundamentals: named users and roles, least privilege, no root usage, and scoped, reviewed third-party access.
• Own secrets and configuration across development, staging, and production, and backups you have tested by restoring.
• Run the AWS account day to day: cost, capacity, alarms, logging, and monitoring across services and vendor integrations.

Quality and delivery

• Write unit, integration, and API tests, including failure paths around external services, and run them in CI/CD.
• Investigate production failures and work out whether the cause is the application, the infrastructure, authentication, data, or a vendor.
• Work with the QA engineer to define what a safe backend release requires.
• Report plainly on what is fragile, what is fixed, and what a proposed change will cost to build. Raise a disagreement once with your reasoning, then commit.

Minimum qualifications

What matters most

For us, these matter most.

• Strong Python skills, with production backends you have written in Python yourself, using Flask, Django, or FastAPI.
• Experience operating an AWS account, not just deploying into one: IAM roles and policies, secrets, alarms you configured, backups you have restored, and access you scoped or revoked without breaking anything.
• Experience writing and owning Terraform, including state and environment separation. Another IaC tool is acceptable if you are ready to switch, but you must have owned it, not inherited it.
• Experience integrating paid vendor APIs in a live user-facing request path and handling them failing: explicit timeouts, capped retries, rate limits, and defined behavior when a provider is slow or down.

Experience

• 4+ years of professional backend engineering, including systems you owned in production.
• Experience designing and consuming REST APIs.
• Working knowledge of AWS beyond the basics: Lambda, EC2, S3, API Gateway, and CloudWatch.
• Solid understanding of asynchronous operations, error handling, retries, timeouts, and rate limiting.
• SQL and relational databases, ideally PostgreSQL: joins, indexes, query plans, and the judgment to question a result that looks too clean.
• Some production exposure to LLM or other AI-powered APIs.
• Experience writing automated tests for backend systems, debugging production applications, and logging and monitoring.
• Experience using AI coding tools such as Claude Code, Cursor, or Copilot, while reviewing and owning what they produce.

Ways of working

• Able to work independently and take ownership of your surface area.
• Excellent written and spoken English. Much of the work is written down, and our users and stakeholders are in North America.
• At least two hours of daily overlap with US Eastern time, and able to be in the Gurugram office 3–4 days a month.

Preferred qualifications

• Experience managing AI API cost, usage, and quotas with OpenAI, Anthropic, Google, AWS Bedrock, or similar.
• Experience with systems depending on several external APIs at once.
• Experience with SQS, EventBridge, DynamoDB, and asynchronous or event-driven architectures.
• Experience with backends serving a native mobile app, including server-side subscription or receipt verification.
• Enough mobile development knowledge to read the app codebase and understand what it is sending.
• Experience on small teams where engineers carry significant ownership.