Required Experience, minimum: 8 years

Required Experience, maximum: 12 years

- Minimum Qualification: B.Tech/B.E

Role

Lead Agentic Engineering & AI Platform Engineer (Expert)

Visit our website bmwtechworks.in to know more.

About the UNIT/ Unit Overview

AIRE (AI-native Requirements Engineering) is a new BMW Group tool for managing vehicle requirements, the descriptions of what a car and its software must do. It keeps the full history of every requirement, shows how requirements are linked to each other and to tests, and shows what is affected when something changes. It helps teams meet automotive standards such as Automotive SPICE and ISO 26262. AIRE is built for AI agents first and a user interface later, and the team uses AI agents to help build it. In this role you set up the AI agents, tools and rules the team uses to build AIRE, and the AI features inside the product.

Location

Pune / Bangalore

Experience

Number of openings

1

What awaits you/ Job Profile

- Design, build and operate the AI-native delivery harness for AIRE across the whole SDLC: requirements, architecture, implementation, testing, documentation, release and operation.

- Own the shared agent assets: agents, skills, instructions, specification templates and MCP tool configurations. Keep them versioned, reviewed, tested and aligned with the architecture and the requirements.

- Define and implement the verification loops for agent-generated work, including static checks, tests, AI evaluations, architecture fitness functions and the required human approval gates.

- Build and run the evaluation pipeline for AI functions in the product and in the harness, together with the Senior Quality & AI Evaluation Engineer.

- Integrate BMW-approved LLM services securely, including identity, secrets, data classification, rate limits,



cost monitoring and logging of prompts and tool calls where permitted.

- Ensure that agent-generated changes are traceable to requirements and to the accountable human approver, to support audit and tool-qualification evidence.

- Measure and report the effectiveness, quality and cost of AI-native delivery to the Lead Value Flow Engineer, and continuously improve the harness.

- Coach the team in agentic practices and contribute reusable assets to BMW-wide AI engineering communities.

What should you bring along

- Degree in Computer Science, Software Engineering, Data Science or a related field, or equivalent relevant experience.

- Minimum of eight years of software or platform engineering experience, including at least two years of hands-on work with LLM-based agents and AI-assisted software delivery.

- Demonstrated track record of introducing developer tooling or platforms that measurably improved delivery.

- Proven ability to work across engineering, security and governance stakeholders without requiring line-management experience.

Must have technical skill

- Proven hands-on experience applying AI-native engineering in real software delivery: coding agents, specification-driven development, multi-agent workflows, and bounded implementation and verification loops.

- Strong experience designing and curating reusable agent assets such as agents, skills, instructions, prompts, specification templates and tool integrations, including their versioning and review.





- Experience with the Model Context Protocol (MCP) and with tool-use patterns for LLM agents, including least-privilege tool access, sandboxing and guarding against prompt injection.

- Experience building evaluation pipelines for LLM-based components and agents: golden datasets, regression evaluations, LLM-as-judge with human calibration, and acceptance thresholds in CI.

- Strong platform engineering skills in Python and TypeScript, CI/CD (for example GitHub Actions), containers and cloud services on Microsoft Azure.

- Understanding of context engineering, retrieval-augmented generation, model selection, rate limits, cost control and observability of LLM usage.

- Ability to define human approval gates and separation of duties in an agentic SDLC, so that every agent-generated change remains reviewable, traceable and attributable.

- Ability to measure the effect of AI-native delivery, for example acceptance rate, rework rate, cycle time, evaluation pass rate and cost per change, and to improve the harness based on these measurements.

- Solid enablement skills to coach engineers in effective, safe and responsible agent use.

Good to have technical skills

- Prior experience with BMW BASE (BMW Agentic Software Engineering), the GATE LLM API and the governance of FG-45 Central AI.

- Knowledge of the EU AI Act, AI risk management and responsible-AI practices in an enterprise context.

- Experience applying AI to requirements engineering tasks such as requirement quality checks, classification, linking suggestions and impact explanation.

- Experience with agent frameworks and SDKs (for example LangGraph, Semantic Kernel, Google ADK or the GitHub Copilot SDK).

- Experience with the qualification of AI-assisted tooling in regulated environments.

📌 Lead Agentic Engineering & AI Platform Engineer (AIRE) (Pune)
🏢 BMW TechWorks India
📍 Pune