As a Senior Software Engineer on the Data Plane engineering team, you will help build Hume’s on-premises inference and workflow platform for running large-scale AI workloads. You will work closely with backend engineers, research scientists, machine learning engineers, product managers, and infrastructure engineers to bring new models and inference capabilities into production.
This is a systems-oriented role focused on reliability, performance, and efficient resource utilization. You will operate with a high degree of autonomy, make sound architectural decisions, and own systems throughout their lifecycle—from initial design and implementation to deployment, observability, and production support.
Design, build, and maintain the data plane software that executes AI models and voice data workflows.
Develop capabilities such as declarative workflow execution, parallel and fan-out processing, inference-engine integrations, tracing, and execution-status reporting.
Integrate new models and runtimes while maintaining consistent performance, reliability, and operational behavior.
Diagnose complex production issues across application, operating-system, container, networking, and hardware boundaries.
Profile and optimize latency, throughput, memory usage, and resource utilization.
Improve the observability, resilience, testability, and operational safety of the platform.
Collaborate with research and product teams to turn emerging model capabilities into reliable production systems.
Write clear technical documentation for the systems and features you build.
What You’ll Bring
Significant professional experience building server-side or systems software.
Qualification
Experience with GPU inferenceExperience with PythonContributions to backend
Preferred
Demonstrated experience designing or building systems with a control-plane/data-plane architecture.
Experience with PyTorch or other machine-learning runtimes.
Experience with GPU inference, CUDA, ONNX, TensorRT, or performance profiling for accelerated workloads.
Experience with Python, or a willingness to learn it. Model execution and integration often require a small amount of supporting Python code.
Contributions to backend, systems, or machine-learning infrastructure open-source projects.
Familiarity with how neural networks are executed, including model graphs, tensor operations, and forward propagation.
Experience building container images and maintaining CI/CD pipelines.
Experience deploying software into on-premises or customer-managed environments.
