Forward Deployed Software Engineer

About the company

• Position listed on behalf of a partner company managing all applications
• Forward-thinking engineering environment combining software development, AI, and product innovation

Responsibilities

• Partner with clients to translate ambiguous needs into requirements, architectures, prototypes, and delivery plans
• Design, develop, and deploy AI-enabled applications, agentic workflows, APIs, developer tools, data pipelines, and full-stack products
• Build AI harness infrastructure: prompts, tools, context management, memory, permissions, runtime controls, observability, feedback mechanisms
• Develop agent loops supporting planning, tool selection, execution, observation, reflection, recovery, termination
• Model workflows using graphs, state machines, and orchestration patterns
• Create evaluation frameworks and testing strategies (task completion, accuracy, reliability, safety, latency, cost)
• Build proofs of concept and transform prototypes into secure, scalable production systems
• Integrate AI solutions with enterprise data sources, cloud platforms, APIs, databases, third-party services
• Establish safeguards, permission boundaries, approval mechanisms, fallback strategies
• Instrument applications for observability and troubleshoot complex issues
• Apply engineering best practices: automated testing, documentation, security, version control, CI/CD

Requirements

• Experience building and deploying production-grade software, ideally AI-powered applications or agentic systems
• Strong skills in Python, TypeScript, or another language used for AI applications
• Understanding of harness engineering: infrastructure, context, tools, permissions, controls, feedback
• Experience with agent loops: planning, tool selection, execution, observation, reflection, recovery, termination
• Familiarity with graph-based orchestration, state machines, multi-step workflows
• Experience designing evaluation systems: benchmarks, regression suites, model-graded evaluations, human review, production monitoring
• Knowledge of prompt design, context management, memory, retrieval, tool use, structured outputs, model selection
• Experience instrumenting AI systems for observability and debugging