On-Prem AI Infrastructure Engineer (Kubernetes + GPUs)

Pursuit Talent Advisory; Austin (TX); On-site; USD 150,000 - 210,000; Full time

Job summary

Pursuit Talent Advisory is partnering with an early growth-stage company building AI systems for industrial environments. The engineer will own on-prem deployment, run reliability on constrained hardware, and ensure repeatable releases onto customer hardware with VPN constraints.

Expect to write application code, contribute to integration with cloud dev/staging, and own CI/CD, GPU inference, and hardened security practices. Work is hands-on and autonomous in a small, AI-native team.

Qualifications

• Strong Kubernetes fundamentals including StatefulSets, storage, networking, ingress, and debugging.
• Infrastructure as code in production (Terraform or equivalent).
• Docker with multi-service builds and registry workflows.
• CI/CD pipelines ownership and maintenance.
• Experience deploying AI workloads on GPUs with private hardware.
• Linux/ Bash proficiency and ability to read/write Python.

Responsibilities

• Own the on-prem deployment path and ensure reliability on single-node and small multi-node setups.
• Own GitOps with declarative reconciliation, image pinning, and rollback boundaries.
• Own cloud dev/staging infrastructure as code and promote through environments.
• Own CI/CD pipelines for multi-service builds and environments that teardown automatically.
• Own GPU inference layer with model/quantization choices and latency targets.
• Improve operability with runbooks, telemetry, and secure separation of components.
• Reduce toil by automating deployment steps and delegating to agents.

Skills

Kubernetes fundamentals

Terraform

Docker

CI/CD ownership

Linux Bash

Python

nvidia-smi readouts

On-prem hardware

Tools

GitOps (Flux/Argo CD)

GPU inference server (vLLM/TGI/TensorRT-LLM)

Job description

Pursuit Talent Advisory is partnering with an early growth-stage company building AI systems for industrial environments. The engineer will own on-prem deployment, run reliability on constrained hardware, and ensure repeatable releases onto customer hardware with VPN constraints.

Expect to write application code, contribute to integration with cloud dev/staging, and own CI/CD, GPU inference, and hardened security practices. Work is hands-on and autonomous in a small, AI-native team.