AI SWE / Agentic Handover Engineer (DevOps Engineer)

Purpose

Deliver AI-assisted handover and due diligence support with added emphasis on automation, repeatability, evidence generation, and reusable workflow packaging across multiple engineering work packages. A central goal of the role is to create repeatable AI-assisted workflows that help assess, document, and transition a cloud software stack derived from OpenStack.

WHAT WILL YOU DO?

• Build reusable AI-assisted workflows for repository intake, code scanning, dependency extraction, build issue triage, and documentation generation.
• Package prompts, retrieval steps, tool integrations, and evaluation logic into repeatable engineering patterns suitable for enterprise delivery.
• Create workflow accelerators that support the analysis and handover of a cloud software stack derived from OpenStack, including service mapping, control-plane understanding, and cross-component dependency views.
• Help define quality gates for AI-assisted outputs, including traceability, explainability, approval points, and human-in-the-loop controls.
• Partner with technical PMO and work-package leads to produce evidence packs that support decisions, risks, and transition readiness views.
• Improve throughput of technical due diligence activities without compromising engineering quality, auditability, or confidentiality constraints. Examples of market tools and workflow frameworks expected
• Agentic workflow frameworks such as LangGraph, OpenAI Agents SDK, AutoGen, CrewAI, LlamaIndex Workflows, or similar orchestration stacks.
• Tool integration and agent control patterns such as memory, tool calling, structured retries, fallback logic, evaluation loops, and human-in-the-loop review stages.
• Support components such as vector databases, retrieval layers, observability tooling, and API/service wrappers that allow workflows to operate across code, documents, and engineering evidence.
• Engineering workflows that can operate across OpenStack-related repositories, CI/CD systems, architecture metadata, test outputs, and service dependency graphs are especially relevant.