Software Engineer, AI - Dunkirk / Buffalo NY

Position Summary

The Software Engineer – AI is a hands-on role integrating data engineering, software development, and AI agent architecture, working closely with platform engineering, data science, security, and product teams. The Software Engineer – AI will contribute to AI agent development and collaborate with cross-functional teams on the implementation of production-grade AI systems. This role supports Engineering leadership by implementing change-controlled solutions, helping maintain project schedules and quality standards, and providing technical support for daily AI-powered operations — including the implementation and improvement of agent frameworks and supporting infrastructure.

Essential Functions

• Contribute to the design, development, maintenance, and deployment of AI agent systems, including one or more of LangGraph StateGraph patterns, CrewAI multi-agent orchestration, LlamaIndex data/agent workflows, or related frameworks.
• Design, build, and maintain FastAPI-based agent servers, including async/await endpoints, streaming responses, and health check endpoints, following established architecture and coding standards.
• Implement comprehensive testing strategies — unit, integration, edge-case, and performance — for agents and services to ensure production reliability.
• Implement and extend guardrails for AI interactions, including input/output validation, safety checks, prompt hardening, policy enforcement, and robust error handling in alignment with organizational security and compliance standards.
• Implement agent observability including per-step traces, tool-call telemetry, cost/latency budgets, and SLO-based alerting using existing monitoring and logging platforms.
• Help design and maintain agent evaluation sets, adversarial tests, regression tests, and monitoring for safety/quality drift.
• Design and build RAG/retrieval pipelines using vector databases such as Chroma, Milvus, Weaviate, and Qdrant, working with data and platform teams on schema and performance.
• Collaborate with SMEs across Platform Engineering, Data Science, Security, Quality, and Product to ensure agent systems meet technical and operational requirements.
• Contribute to the creation, modification, and maintenance of AI system documentation, including architecture decision records (ADRs), runbooks, API specifications, network topology diagrams, data flow diagrams, and best-practice guides.
• Provide technical input regarding operability, technical feasibility, engineering design, security posture, maintainability, and documentation requirements.
• Assist with the planning and implementation of AI agent-based projects and solutions in collaboration with cross-functional teams, helping deliver on cost, timeline, and quality targets.
• Implement and support containerized deployment, start-up, commissioning, and release qualification activities using Docker multi-stage builds, dependency management, and container health check best practices.
• Collaborate with and learn from senior engineers; may occasionally assist less experienced engineers through code reviews and pairing on troubleshooting AI agent and service-related issues.
• Follow and improve Standard Operating Procedures (SOPs), process improvements, and standard engineering templates for agent development.
• Monitor agent performance and reliability, collect metrics and logs, and use data to propose and implement optimization opportunities.

Education & Experience

• Bachelor’s degree in Computer science, Software Engineering, or a related field with 3+ years of relevant software development experience is required.
• 2+ years of hands-on Python development experience, with proficiency in Python 3.12+ and modern language features (type hints, async/await, etc.) is required.
• 2+ years of hands-on Typescript development experience.
• Experience with at least one AI agent framework (LangGraph, LangChain, CrewAI, or LlamaIndex) in a production or near-production setting is required.
• Experience building and running containerized applications with Docker, including multi-stage builds and foundational DevOps practices, is required.
• Experience with troubleshooting and system optimization in an enterprise or regulated environment is preferred.
• Experience working within an organization with formal change control, compliance, or audit requirements is preferred.