AI Engineer

The client is a global provider of professional software for the health, tax & accounting, finance, legal, and compliance sectors. We're looking for an AI Engineer to design, build, evaluate, and ship enterprise Generative AI and agentic systems that go into production and get used. This is a high responsibility role. You'll own solutions end-to-end for a $7B+ company

NOTE: We read a lot of AI-polished applications. Yes, AI is great (even this JD was made with the help of AI). However, what counts here is sitting across the table and talking through your own work, in your own words. We interview in person and we go deep. Built and understood real things? You'll enjoy it. Leaned on AI to speak for you? It won't hold up.

We believe in fair compensation: if you can walk the talk, you'll be rewarded.

What you'll do:

• Design and develop enterprise AI applications using LLMs and modern AI frameworks.
• Build RAG, conversational AI, single-agent, multi-agent, and agentic workflow solutions.
• Create agents, reusable Agent Skills, custom instructions, prompts, tools, and knowledge integrations.
• Perform agent harness engineering, including orchestration, context, memory, state, tool permissions, retries, human approvals, and error recovery.
• Develop agents using AWS Bedrock/AgentCore, Microsoft Copilot Studio, Azure OpenAI, OpenAI, Anthropic, or equivalent platforms.
• Build orchestration and automation using LangGraph, Strands Agents, CrewAI, AutoGen, OpenAI Agents SDK, LangChain, n8n, Power Automate, or similar tools.
• Design and integrate APIs, MCP servers, connectors, enterprise tools, and knowledge sources.
• Use GitHub Copilot and OpenAI Codex for coding, debugging, testing, refactoring, code review, and documentation.
• Apply prompt engineering, context engineering, model routing, caching, and token optimization.
• Implement evaluation datasets, regression tests, guardrails, adversarial testing, hallucination mitigation, and responsible-AI controls.
• Develop scalable APIs and backend services using Python and FastAPI.
• Implement CI/CD, observability, tracing, monitoring, rollback, and AgentOps practices.
• Apply security-by-design, including least privilege, secrets management, data protection, and safe tool execution.
• Collaborate with business and technical teams to convert requirements into measurable AI solutions.

You're a strong fit if…

• You're AI-native — you already build with LLMs and agents daily, and you know their limits, failure modes, and costs.
• You've shipped things to production and can point to what you owned and what broke.
• You know when not to reach for an LLM, and will use classical ML, statistics, or plain software when that's the better call.
• You obsess over details and don't trust AI output you haven't verified yourself.
• You're a strong core engineer, not just a framework user.

You're probably NOT a fit if…

• You need constant direction before acting.
• You let AI generate the work and sign off on it without going deep yourself.
• Your experience is only coursework or demos that never went to production.

Skills:

• Strong programming skills in Python and experience with FastAPI or similar frameworks.
• Hands-on experience with Generative AI, LLMs, prompt engineering, and context engineering.
• Experience building RAG, agentic AI, multi-agent, and tool-using applications.
• Experience creating agents, Agent Skills, custom instructions, agent tools, and reusable workflows.
• Understanding of agent harness engineering, including orchestration, sessions, state, memory, context management, human-in-the-loop, and recovery.
• Experience with AWS Bedrock/AgentCore, Microsoft Copilot Studio, Azure OpenAI, OpenAI, Anthropic, or equivalent platforms.
• Experience with LangGraph, Strands Agents, CrewAI, AutoGen, OpenAI Agents SDK, LangChain, or similar frameworks.
• Experience with workflow-automation tools such as n8n, Power Automate, or Logic Apps.
• Experience with GitHub Copilot, OpenAI Codex, or equivalent AI-assisted development tools.
• Knowledge of MCP, REST APIs, tool calling, connectors, and enterprise integrations.
• Knowledge of vector databases, embeddings, semantic search, reranking, and RAG evaluation.
• Experience with AI evaluation, guardrails, security testing, observability, and AgentOps.
• Understanding of token optimization, model selection, caching, performance, and AI cost management.
• Familiarity with Git, Docker, cloud platforms, infrastructure as code, and CI/CD.
• Strong analytical, debugging, communication, and problem-solving skills.

• Experience:

• 1–3 years of hands-on experience building and shipping real AI/software systems in production. This is not a fresher role; you should have substantive, demonstrable work behind you (shipped systems, real users, real data).