Software Engineer - IV

The role in a gist: As FlexiLoans enters the next orbit of growth, we are looking for people with the passion and hunger to make a dent in the MSME ecosystem. In this role you will be AI-first in how you build. AI coding agents are a normal part of our workflow: writing, reviewing, testing and refactoring code, and understanding unfamiliar parts of a large system. What matters is not tool familiarity by itself, but that what you ship is reviewed, tested and production-ready, at the pace this way of working makes possible. You will also work across the stack. Backend and frontend are both part of the role, and we would like you to be comfortable on either side of the API boundary.

What you will work on:

Backend

• Strong JavaScript/TypeScript and Node.js. You can design an API, not only implement one.
• Relational data modelling and SQL you can reason about indexes, query plans, transactions, migrations on live tables.
• Vendor and third-party integrations, done well: authentication, retries, idempotency, webhooks, rate limits, sandbox-to-production rollout, and reconciliation when the other side fails mid-flow. Comfort working directly with a vendor's technical team to get an integration over the line.
• Queues, background jobs and scheduled work, including what happens when a job runs twice.

Frontend

• React with TypeScript. Component design, state management, forms that hold up under real user behaviour.
• Enough CSS and layout sense to build a clean screen without a pixel-perfect design handed to you.
• Performance and correctness on the browser side: loading states, error states, the ugly edge cases.

Engineering craft

• Tests you actually write. Shipping agent-assisted code at volume only works if the test coverage holds it up.
• Git discipline, CI/CD, and enough DevOps grounding to be self-sufficient: containers, environment configuration, pipelines, deployments and rollbacks. You do not need to be an SRE, but you should not need one for routine work.
• AWS fundamentals: how your service runs, scales and fails.

Monitoring and cost awareness

• Instrumenting what you build logs, metrics, alerts so problems surface on a dashboard rather than in a customer complaint.
• Reading monitoring output and acting on it: triaging an alert, finding the root cause, and closing the loop with a fix or a tuned threshold rather than muting it.
• An eye on infrastructure cost. Knowing roughly what your service costs to run, spotting the query, log stream or API call pattern that is quietly expensive, and factoring that into design decisions.

How you operate

• You are comfortable with ambiguity and will go and resolve it often by asking the ops or product team directly.
• You share status without being prompted, including when the news is not good.
• You care whether the change worked for the business, not only whether the story was closed.

What you will own

You own outcomes, not tickets. Concretely, for the problems assigned to you:

• Understand the problem first-hand. Spend time with the product owner or the ops team that has the problem, and see how the work happens today rather than working from a requirements document alone.
• Design, build, and ship end to end. API and data model through to the screen the user touches. Backend and frontend are both yours.
• Take it to production and keep it healthy. Deploy it, monitor it, and act on what the monitoring tells you. Errors, latency, failed jobs, and data quality are yours to watch, and ideally you spot the issue before someone reports it.
• Be mindful of what it costs to run. Every service you deploy carries an infrastructure bill. Think about instance sizing, storage, query patterns, log volume, and the cost of third-party API calls before you ship, and raise it early when something is likely to get expensive.
• Integrate cleanly with vendors and partners. A large part of our platform is third-party: bureaus, banks, KYC providers, payment rails, co-lenders. Working well with vendor teams, reading their documentation carefully, and handling their failure modes gracefully is core to the job, not a distraction from it.
• Be accountable for what you commit to. Give estimates you believe in, flag slippage early, and keep the status visible.
• Leave it documented. A runbook, an API contract, and a short note on any non-obvious decision, so the next engineer or the next agent can pick it up without you.
• Build securely by default. We are a regulated lender. Customer and credit data handling, access control, audit trails, and PII discipline are part of the definition of done, not a later hardening pass

Qualification and Experience:

• 7+ yrs. A domain spanning services and teams. Sets the plan; trusted on the hardest and riskiest work. Owns architecture and the trade-offs behind it. Partners with function heads; helps decide what not to build
• BS/BE/BTech degree in Computer Science or Engineering.