Python Developer | Wilmington/New York, US

Core Responsibilities 

• Build and operate production agentic AI systems, including deployed, monitored, and debugged agents running at scale. 

• Own agentic use cases end-to-end, partnering directly with business stakeholders from problem definition through delivery and operational support. 

• Design orchestration patterns for autonomous or semi-autonomous agents using modern agent frameworks and production orchestration layers. 

• Develop enterprise capabilities as tools or skills for LLMs, enabling agents to interact with business systems and workflows in a controlled, scalable way. 

• Engineer reliable backend services for agentic workloads, with emphasis on Python-based development and integration with Java service layers where required. 

• Implement and support production infrastructure for AI workloads, including environments where Kubernetes and model-serving components such as vLLM may be part of the stack. 

• Evaluate and communicate system failure modes, including the ability to walk through shipped systems, operational issues, debugging approaches, and mitigation strategies. 

• Collaborate with vendor, engineering, and business teams to deliver solutions with limited ramp-up time, consistent with expectations for senior contract engineering talent. 

• Maintain a production-first engineering standard, ensuring the role does not over-index on framework familiarity at the expense of real deployment experience. 

Required Qualifications 

• 7+ years of software engineering experience, with demonstrated experience building and shipping production systems. 

• Hands-on production experience with agentic AI or GenAI applications, including deployment, monitoring, debugging, and operating agents at scale. 

• Comfortable in a senior contract delivery model where the expectation is faster delivery and limited ramp-up investment. 

• Core expectations (must have exposure across all): 

• Software engineering: Proficiency in Python (preferred) or Java, with solid fundamentals (clean coding, testing, APIs, debugging, performance, and maintainability). 

• Cloud engineering: Working knowledge of at least one public cloud provider—AWS preferred—including hands-on familiarity with deploying/running services, IAM/security basics, logging/monitoring, and common managed services. 

• AI/ML exposure: Practical exposure to AI/ML concepts and workflows (e.g., model integration patterns, basic ML lifecycle understanding, evaluation basics), sufficient to ramp into agent development with minimal hand-holding. 

• Preferred experience (strong differentiators): 

• Hands-on experience integrating with frontier model APIs (e.g., OpenAI or similar), including prompt design, tool/function calling, streaming, and reliability/cost considerations. 

• Familiarity with advanced agent architecture concepts such as memory, guardrails/assurance, and MCP (Model Context Protocol) or similar patterns for tool interoperability. 

• Experience with conversational/agent frameworks or platforms, such as LangGraph, CrewAI, Rasa, Decagon, or equivalent orchestration frameworks, including building multi-step workflows and tool-based agents 

Preferred Qualifications 

• Experience with LangGraph as a production orchestration layer. 

• Experience with vLLM or comparable model-serving infrastructure. 

• Experience in regulated-industry or financial-services technology environments, especially where enterprise scale and production-path stakes are important. 

• Experience working with business stakeholders to deliver end-to-end AI use cases, not only platform or prototype work. 

• Ability to operate in an onsite or hybrid delivery model, especially in a market such as NYC where the attachment notes a deeper finance-AI contractor pool. 

• Position the role as a senior contract role with a production-delivery bar, not as an entry-level GenAI experimentation role. 

• A 12-month engagement is reasonable for senior contractors, though onsite expectations may narrow the candidate pool. 

• If submissions are limited, relax requirements in this order: Java first, onsite days second, and do not relax the production-experience bar. 

Suggested Must-Have Screening Criteria 

• Can clearly describe at least one agentic AI system they personally helped ship to production. 

• Can explain failure modes, monitoring, debugging, and operational lessons learned from a shipped system. 

• Has strong Python capability and either strong Java experience or willingness to work daily with a Java service layer. 

• Has practical experience with LangChain, LangGraph, Kubernetes, or comparable production AI infrastructure. 

• Demonstrates ability to own use cases end-to-end with business stakeholders. 

Recommended Positioning Language for the Job Description