Job Description & Details
This is a heavy-hitting engineering gig bridging the gap between traditional enterprise software development and cutting-edge generative AI. You'll be building intelligent employee experiences for a major financial institution, combining deterministic workflows with autonomous agentic systems. If you like playing with LLMs, orchestrating complex APIs, and actually deploying AI into production rather than just writing prompt demos, this one is worth your time.
What You'll Actually Be Doing
Expect to spend your days designing and deploying virtual assistants, conversational AI, and agentic workflows that integrate deeply with enterprise tech stacks. You won't just be writing prompts; you'll build the underlying orchestration components, Model Context Protocol integrations, and deterministic business process automations that connect multiple disparate systems together. You'll also be on the hook for performance tuning, observability, and making sure everything complies with strict banking security and governance frameworks.
The Core Tech Stack
You'll need a rock-solid foundation in either Java or Python, as you'll be writing actual production backend code rather than just gluing scripts together. Deep hands-on familiarity with Large Language Models, prompt engineering frameworks, retrieval systems, tool calling, and API integrations are non-negotiable since you'll be building hybrid architectures that blend traditional software logic with AI reasoning.
Interview Expectations
Expect the hiring team to throw a real-world scenario at you, such as asking how you would design a hybrid workflow where a deterministic business process needs to hand off execution to an autonomous AI agent, and how you would handle failure states if the LLM hallucinates or times out. They want to see that you understand the reliability limits of generative AI and know how to build fallback mechanisms using robust software engineering patterns. You'll also likely get grilled on API design and security, specifically regarding how you safely expose enterprise databases and tools to an LLM via Model Context Protocol.
Application Advice
Skip the generic AI buzzwords on your resume and focus heavily on your actual engineering chops with Java or Python alongside concrete examples of LLM orchestration or conversational AI deployments. Make sure to highlight terms like 'API integration,' 'agentic workflows,' 'prompt frameworks,' and 'observability,' as the automated screening systems at a major financial institution are specifically hunting for these operational keywords.