SDE III (Backend) Agentic SDLC

The candidate will have responsibilities across the following functions:

Backend engineering:

• Design, build, and own backend services in Java, Spring Boot, and a microservices architecture with real accountability for performance, scalability, and robustness.
• Own server-side logic, data models, APIs, and integrations end-to-end.
• Drive HLD and LLD for new services and for material refactors of existing ones.
• Agentic SDLC ownership.
• Own how agentic tooling is applied across our SDLC spec/design, implementation, review, testing, and production monitoring, not just at the coding step.
• Build and maintain the scaffolding that makes agents effective on a large codebase: repo-level context and instruction files, custom agents/subagents, slash commands and reusable workflows, MCP integrations to internal systems (Jira, BigQuery, observability, docs).
• Define the review bar for agent-generated code: what gets human-reviewed, what gets gated by tests, what never gets delegated.
• Instrument and evaluate the workflow cycle time, review turnaround, escape defect rate, and test coverage on agent-authored changes and iterate based on that data, not vibes.
• Raise the team's ceiling: onboard engineers onto these workflows, run internal enablement, and set guardrails for security, licensing, and data handling when agents touch source code and production data.
• Leadership: Lead across teams cross-functional design reviews, technical direction, and mentoring senior and mid-level engineers.
• Make and defend build/buy/delegate decisions on tooling.

Requirements:

• 6+ years in backend engineering, with at least 2 in a lead or tech-lead capacity.
• Strong proficiency in Java, Spring Boot, Hibernate/JPA, and microservices.
• Demonstrated experience in HLD and LLD, and in designing, building, and deploying microservices-based systems in production.
• Hands-on agentic SDLC experience within the last 12 months: you have shipped production software where AI agents were a primary part of the workflow. Concretely, experience with tools such as Claude Code, Cursor, Codex, Devin, Copilot Workspace/agent mode, Aider, or equivalent, applied to at least three of: design, implementation, code review, test generation, and production debugging/monitoring.
• Practical judgment about where agents fail in context management on large codebases, hallucinated APIs, silently wrong tests, review fatigue and concrete mitigations you've put in place.
• Solid grounding in Git, CI/CD, and automated testing, including how these change when a large share of dis are agent-authored.
• Strong SDLC fundamentals and a track record of working with multiple teams.

Nice to have:

• Built custom agents, subagents, or MCP servers against internal systems.
• Prompt/context engineering at the repo scale (e. g., CLAUDE. md-style instruction files, retrieval over internal docs, codebase indexing).
• Experience with LLM evaluation, regression harnesses, or accuracy pipelines.
• Observability tooling (New Relic, Datadog, Prometheus/Grafana) and agent-assisted incident triage.
• Healthcare, FHIR/HL7 or medical coding domain exposure.
• Python for tooling and data work; PostgreSQL, Elasticsearch, or Neo4j; GCP or AWS.