AI Platform Engineer — Agentic SDLC

• Extend and harden an AI-driven software development lifecycle platform built on Kiro IDE. The platform uses autonomous agents, MCP tool integrations (Rally, Playwright, Git, Confluence), steering files, and event-driven hooks to orchestrate requirements gathering, architecture design, code generation, test automation, and deployment gating — all AI-first.

  build new agent capabilities, design RAG pipelines for context-aware code generation, create governance guardrails, and integrate with enterprise ALM/DevOps tools. This is not a testing role or a DevOps role — it's building the platform that makes all SDLC phases AI-native.

Key Areas of Responsibilities:

• Design and build AI agents that orchestrate multi-step SDLC workflows (fetch → generate → execute → validate → sync)
• Build RAG pipelines that feed architecture docs, coding standards, and Rally artifacts as grounded context to LLMs
• Implement MCP servers and tool integrations (Rally, Playwright, Git, Confluence, Azure DevOps)
• Author steering files (persona, governance, lifecycle) that constrain agent behavior with formal rules
• Build self-healing and auto-remediation systems with bounded autonomy and human escalation
• Design state machines for artifact lifecycles (requirements → design → code → test → deploy)
• Create event-driven hooks for CI/CD integration, nightly runs, and release gating
• Implement property-based testing to formally verify system correctness
• Package and deploy agent infrastructure (Docker, pipelines, credential management)

Skills Required:

Core Language

TypeScript / Node.js

AI/LLM

Azure OpenAI (GPT-4), prompt engineering, structured outputs

Agent Framework

Kiro IDE agents, MCP (Model Context Protocol), tool-use patterns

RAG / Context

Embedding pipelines, vector search, document chunking, context grounding

Testing

Playwright, fast-check (PBT), Jest

ALM Integration

Rally, Azure DevOps, Jira (bidirectional sync)

DevOps

GitHub Actions / Azure DevOps Pipelines, Docker, AKS

Frontend

React 18, TailwindCSS, shadcn/ui (for dashboards/reporting)

Backend

Express, Prisma, PostgreSQL, Redis/Bull queues

Documentation

Markdown-first, Word export, Mermaid diagrams

Must-Have Skills

• AI Agent Systems — Built or extended multi-step autonomous agents with tool-use, guardrails, and human-in-the-loop patterns
• RAG Pipelines — Designed retrieval-augmented generation systems (embedding, chunking, vector DB, context injection)
• TypeScript / Node.js — Production-grade, 4+ years
• LLM Integration — Azure OpenAI or equivalent; prompt engineering, structured outputs, function calling
• Full-Stack Architecture — Can design end-to-end systems (API, queue, DB, frontend, infra)
• CI/CD & DevOps — Pipeline authoring, Docker, scheduled automation, release gates

State Machine / Workflow Design — Lifecycle management, valid-transition enforcement, event-driven orchestration

Nice-to-Have

• MCP (Model Context Protocol) experience or similar tool-use protocols
• Playwright test automation
• Rally / Azure DevOps API integration
• Property-based testing (fast-check, QuickCheck)
• Salesforce application
• Experience with Kiro, Cursor, Windsurf, or similar AI-native IDEs
• Formal verification or correctness-by-construction approaches

Backfill position for Nayeem Mohammed (Emp ID:700971) in GE digital under Manish Purwar