Software Engineer - AI Agent Development
Role summary
Seeking an experienced contract Software Engineer to design and build AI agents for device onboarding and environment monitoring in data center infrastructure. This role focuses on applied AI engineering, requiring the development of production-grade agents on Microsoft Azure AI Foundry. Key responsibilities include automating device discovery and registration, and ensuring the platform's digital representation of the physical environment remains accurate and timely. The agents must communicate directly with physical devices using protocols like Modbus, BACnet, and OPC-UA, and be built with extensibility and maintainability in mind. The initial engagement is for 3 months with potential for extension.
Job Description
Software Engineer - AI Agent Development
Overview
seeking an experienced contract software engineer to designand build AI agents that support device onboarding and continuous environment monitoring fordata center infrastructure. The engagement is an initial 3-month term with the possibility ofextension based on delivery.
The agents will be built on Microsoft Azure AI Foundry and must be production-grade from thestart. This is applied AI engineering work, not research. The contractor will be expected todeliver agents that operate reliably against real device data, integrate cleanly with existingplatform services, and are structured so that other engineers can extend and maintain them.
What You Will Build
- The first area of work is automated device onboarding. The agents will handle the discovery andregistration of new devices into the platform, reducing the manual effort currently required tobring physical equipment under management. This includes identifying device types, extractingconfiguration, and populating the platform with accurate asset records without requiring hand-holding from operations staff.
- The second area is continuous environment monitoring. The agents will detect changes to thephysical data center environment and keep the platform's representation of that environmentconsistent with physical reality. When equipment is added, removed, reconfigured, or behavesunexpectedly, the agents are responsible for surfacing that and driving reconciliation. Accuracyand timeliness of the digital representation is the success metric.
- Both workstreams require the agents to communicate directly with physical devices using theprotocols those devices speak.
Required Quali
fi
cations
- 8 or more years of professional software engineering experience, with a demonstratedrecord of delivering production systems.
- Deep expertise in software design and architecture. This role involves substantial use of AI-assisted development tooling, and success depends on the ability to produce precise,unambiguous specifications that can be implemented reliably and verified against clearacceptance criteria.
- Hands-on experience building AI agents or agentic workflows in production environments,including tool use, state management, and reliable execution under real-world conditions.
- Practical experience with Microsoft Azure AI Foundry or a directly comparable enterpriseAI platform for building and deploying agent-based services.
- Strong Python development skills. Agent services will be built primarily in Python andmust be well-structured for extensibility and long-term maintainability.
- Demonstrated experience working with industrial device protocols, specifically Modbus,BACnet, and OPC-UA. The contractor must be able to build agent integrations thatcommunicate with physical devices accurately and reliably.
- Experience working with digital twin concepts or asset registry systems, including thechallenges of keeping a software representation of a physical environment consistent overtime.
Preferred Quali
fi
cations
- Experience with additional OT protocols such as MQTT/Sparkplug-B or DNP3.
- Background in data center infrastructure, facilities management systems, or DCIMplatforms.
- Familiarity with automated device discovery and configuration extraction acrossheterogeneous hardware environments.
- Experience designing agentic systems that operate with human-in-the-loop approvalworkflows in safety-sensitive or operationally critical contexts.
