Software Engineer - LLM Infrastructure

Swoop Technologies Job PostingSwoop Technologies has a mission to organize and make accessible the world's military and critical infrastructure. We are building a distributed operating system, SwoopOS, that decomposes the world's equipment into a distributed robotic embodiment upon which a new generation of distributed systems, autonomous systems, and agentic AI can be built and deployed using our SDK, Valhalla, and operated via our browser, Surf. Imagine the world's equipment - consisting of the electrical grid, communications architectures, manufacturing facilities, and militaries as a trapped supply of inputs possessing the potential to ensure Western military advantage, sovereign control of economically competitive manufacturing capacity, or the creation of a grid that fosters energy dominance. Swoop is liberating these trapped assets, allowing them to contribute to the world's future as a series of building blocks to be combined at the speed of software, limited by only the hard constraints of physics and the soft constraints of safety. That is what Swoop is building. Not in the data center or cloud or edge on-premise computing node. In the physical world. This is a hybrid position that requires someone based in Minneapolis/St. Paul OR Washington, DC who can work in-office 3+ days per week.Swoop's operating system challenges many paradigms defining a scalable API. Management, security, and interoperability each will be re-imagined in how Swoop OS's expose interfaces to inputs and outputs within the stack. In this role, you will develop, maintain, and scale Swoop's self-hosted, custom-tuned LLM. A key objective is advancing how the model understands and represents complex system interactions, providing context-aware insights into system behavior, dependencies, and operational dynamics. You will enable users to ask natural, unconstrained questions about their infrastructure and systems, receiving precise, insight-driven responses. The ultimate goal is to accelerate decision making and drive greater autonomy across distributed infrastructure powered by Swoop OS.What You'll Do:Develop and maintain Swoop's LLM offeringExpand the capabilities of the LLM to interact with the system via tool callsExpand the data searching capabilities of the LLMWork hand-in-hand with frontend developers to build out new LLM features and improve existing onesMonitor the resource usage of installations and make sure that the LLM offering is as efficient and fast as possible given the inference hardware availableMaintain and optimize inference engine architectureTune data storage configurations to optimize for scale and near real-time availability in a streaming architectureEnsure our services have strong availability and service level agreements across our code base, especially as it pertains to the runtime of our Kubernetes cluster in productionYou Should Have:Bachelor's degree in Computer Science or related technical field, or equivalent technical experienceFirm understanding of scalable large language model infrastructureExperience with low-level NVIDIA drivers and NVIDIA Kubernetes Container ToolkitFamiliarity with designing RAG information retrieval systems and time-series anomaly detectionExperience with PyTorch, training and fine-tuning Machine Learning models for resource-light environmentsExperience with Kubernetes in a production environmentProficient Python coding ability with good understanding of data structures and data modelsActive US Security clearance or ability and willingness to be sponsored for a US Security clearanceBonus if you have:Experience with on premise or self-hosted AIExperience with numerous GPUs and understanding of performance characteristicsExperience standing up inference engines such as vLLMSwoop Technologies is proud to be an equal opportunity employer. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, age, national origin, disability, protected veteran status, gender identity or any other factor protected by applicable federal, state, or local laws.