About the Role

We are looking for a Software Engineer (SDE-II) with strong Applied AI fundamentals and an AI-Native engineering mindset to join Flipkart's Fintech & Payments Engineering Group (FPG).

FPG powers payments, credit, and fraud infrastructure for hundreds of millions of Indians. The complexity, scale, and regulatory constraints of this domain make it one of the most demanding environments in which to build autonomous engineering agents. You will work at the frontier of what agentic systems can do, an environment where correctness failures have direct financial and trust consequences..

You will grow rapidly alongside some of the sharpest fintech and AI engineers in the country, solving problems at a scale.

Job Responsibilities

1. Agentic Development Lifecycle (ADLC)

- Design and build autonomous agent pipelines that cover end-to-end ADLC stages: requirements analysis, technical design, code generation, code review, test generation, regression validation, deployment sequencing, and post-deployment monitoring.
- Architect multi-agent systems for engineering workflows: orchestrator agents that decompose engineering tasks, specialist sub-agents (design agent, coding agent, test agent, deployment agent), and critic agents that validate agent outputs before human review.
- Build human-in-the-loop checkpoints into agentic pipelines — define precisely where agents operate autonomously, where they escalate for human validation, and what constitutes a safe boundary for unattended agent execution in a regulated financial setting.
- Implement agent memory and context management for long-horizon development tasks: persistent agent state across ADLC stages, codebase context injection, retrieval-augmented design and code generation.
- Build agent tool suites for software development tasks: code execution sandboxes, static analysis integrations, test harness runners, schema validators, API contract checkers, and deployment pipeline hooks that agents invoke autonomously.
- Design agent evaluation frameworks: measure agent output quality, code correctness, design adherence, test coverage, and regression rates — across development task types and domain complexity levels.
- Operate production agentic pipelines: monitor agent reliability, detect agent failures and hallucinations, manage agent versioning, and continuously improve agent performance through feedback loops and fine-tuning signals.

2. Domain-Specific Agent Engineering for Fintech

- Build agents that understand the fintech and payments domain: agents that reason about idempotency, exactly-once semantics, transaction integrity, PCI-DSS compliance constraints, and financial audit requirements when generating designs and code.




- Design automated validation agents that check agent-generated payment code against domain correctness criteria: settlement logic, reconciliation invariants, fraud signal handling, and regulatory compliance rules.
- Build test-generation agents specialised for financial workflows: agents that generate edge-case scenarios for payment failures, partial settlements, currency rounding, duplicate transaction handling, and idempotency violations.
- Design deployment and incident response agents: agents that autonomously execute canary deployments, monitor payment success rate SLOs, detect regressions, and trigger rollbacks without human intervention.

3. Agent Infrastructure & Platform

- Build the shared agent infrastructure layer for FPG: agent orchestration framework, tool registry, shared memory stores, evaluation pipelines, and agent observability — enabling multiple teams to deploy ADLC agents on a common platform.
- Design agent communication protocols: how agents hand off tasks, share intermediate outputs, resolve conflicts between agent decisions, and maintain consistency across a multi-agent development pipeline.
- Implement agent safety and guardrails: rate limiting on agent-initiated actions, human approval workflows for high-risk operations (production deployments, schema migrations, third-party API changes), and audit logging of all agent decisions.

4. Technical Leadership

- Define the ADLC architecture and agent design standards for FPG — establish what good agent system design looks like across reliability, correctness, auditability, and domain-appropriateness.
- Mentor SDE-I and SDE-II engineers on agent system design, evaluation, and operation — raise the team's capability to build and operate autonomous development pipelines.
- Drive the adoption and evolution of the ADLC across FPG teams — measure adoption, track ADLC-driven productivity improvements, and identify the next stages of the development lifecycle to automate.

What Success Looks Like

- Autonomous agent pipelines you build measurably accelerate FPG's development velocity — shipping well-tested, domain-correct fintech features faster with fewer human engineering cycles.
- Agent systems operate reliably in production: low hallucination rates on domain-specific tasks, correct escalation to humans at appropriate checkpoints,



and verifiable correctness on regulated financial code.
- The ADLC platform you build becomes the foundation other FPG teams build their own domain-specific development agents on.
- FPG engineers spend more time on high-judgment architectural and product decisions.

You (Must Haves)

- 8+ years of software engineering experience, with hands-on production experience building autonomous or semi-autonomous agent systems.
- Deep experience with multi-agent orchestration: task decomposition, agent-to-agent handoffs, critic/validator agent patterns, and managing agent state across long-horizon tasks.
- Experience building agents with real tool use: code execution, file system operations, API calls, database queries, test harness runners and agents that take consequential actions.
- Strong understanding of agent reliability engineering: failure modes (hallucination, tool misuse, context loss, runaway loops), detection strategies, and guardrail design.
- Experience designing human-in-the-loop systems: defining autonomous vs. supervised boundaries, building approval workflows, and audit-logging agent decisions for compliance.
- Strong distributed engineering fundamentals: event-driven architectures, APIs, message brokers, distributed caches and databases.
- Proficiency in Java and Python. Deep familiarity with LLM APIs and agent frameworks
- BE/BTech in Computer Science or equivalent.

Nice to Haves

- Experience in fintech, payments, or financial services — understanding the correctness, compliance, and reliability standards that agent-generated code must meet in this domain.
- Experience building code-generating or code-reviewing agents: agents that write, refactor, or validate production code with measurable quality outcomes.
- Familiarity with LLM fine-tuning, reinforcement learning, or preference optimisation for improving domain-specific agent task performance.
- Experience with MCP (Model Context Protocol) server development — building tool surfaces that agents consume.
- Published work on agentic systems, ADLC methodologies, or AI-driven software engineering.

At Flipkart, we're driven by our purpose of empowering every Indian's dream by delivering value through innovation in technology and commerce. With a customer base of over 350 million, product coverage of over 150 million across 80+ categories, a focus on generating direct and indirect employment and a commitment to empowering generations of entrepreneurs and MSMEs, all driven by a sustainable growth strategy – Flipkart is maximizing for customers, stakeholders, and the planet at large!

Flipkart is a part of the Walmart-owned Flipkart Group, which also includes group companies Flipkart Health+, Myntra, and Cleartrip.

📌 FPG Engineering (Bengaluru)
🏢 Flipkart Internet
📍 Bengaluru