Senior Software Engineer

JLL empowers you to shape a brighter way.  

Our people at JLL are shaping the future of real estate for a better world by combining world class services, advisory and technology for our clients. We are committed to hiring the best, most talented people and empowering them to thrive, grow meaningful careers and to find a place where they belong.  Whether you’ve got deep experience in commercial real estate, skilled trades or technology, or you’re looking to apply your relevant experience to a new industry, join our team as we help shape a brighter way forward.   

About the Role

As a Staff Software Engineer on the MarTech Engineering Team, you'll be a forward deployed engineer embedded in the marketing organization, building production AI agents that solve real campaign, content, social, ABM, and performance problems. You'll work shoulder-to-shoulder with marketing domain experts (SMEs) — sitting in their workflows, learning their systems, and turning their expertise into agents that go live and stay live.

Your remit spans the full lifecycle: rapid prototyping with SMEs to validate ideas, building the integrations and skills needed to ship a working pilot, and hardening pilots into production-grade agents that operate reliably at scale. When something is missing — a CRM that doesn't expose what you need, a data source not yet in the warehouse, a guardrail that doesn't exist — you'll either build a pragmatic workaround to unblock the pilot or translate the gap into a clear, prioritized requirement for our partner platform team — the engineers who own the underlying marketing systems, data plumbing, and AI-native tooling — to address. You sit at the seam between domain expertise and platform infrastructure, and your job is to make sure neither side waits on the other.

The role demands deep technical judgment: when to build vs. wait, when to abstract a one-off solution into a reusable skill, and how to ensure agent outputs translate into safe, brand-aligned, and factually correct actions across a complex marketing stack. Success is measured by the volume and quality of marketing work that agents are actually doing in production — not by demos, not by POCs, but by agents that marketers trust to do the job. This is a fantastic opportunity to build the agentic layer of one of the world's largest commercial real estate platforms from the ground up, and to see your work change how marketing operates at global scale.

Who You Are

We're optimizing for three things ahead of tool-specific experience: deep technical fundamentals, a pragmatic problem-solving disposition, and exceptional cross-audience communication. The agent tooling landscape changes every few months — what we can't easily teach is how someone thinks, ships, and explains under uncertainty. If you have those three, we'll happily back you while you skill up on whichever runtime, framework, or model provider this quarter favors.

• You have a Bachelor's or Master's degree in Computer Science, Engineering, or a related field, or equivalent work experience
• You are proficient in English, both written and verbal, sufficient for success in a remote and largely asynchronous work environment
• You have 5+ years of software engineering experience working in production systems, including significant time integrating against enterprise APIs (CRMs, CMSes, DAMs, ad platforms, marketing automation tools, analytics)
• You have hands-on experience with modern LLM APIs across providers — including prompt engineering, tool use / function calling, structured outputs, and context engineering — and you've worked with enough of them to know that the underlying patterns transfer even as the specific APIs evolve
• You have experience designing agent systems: multi-step reasoning, tool orchestration, memory, error recovery, and human-in-the-loop escalation paths
• You've worked through the hard parts of agent engineering firsthand — hallucination grounding, tool reliability and silent failures, evals that actually predict real-world behavior (and the gap when they don't), cost and latency tradeoffs, prompt drift, and the difference between "works in a demo" and "still works on day 30." You can talk concretely about what you've tried, what's broken on you, and what you've learned. We care more about depth of engagement with these problems than years logged
• You have experience building RAG systems — embeddings, vector stores, retrieval optimization, and grounding — and a strong intuition for when retrieval is the right answer vs. when a tool call, fine-tune, or schema chang