Zilker Partners
Principal Backend Engineer - Kubernetes & Inference Pipelines
Company Overview
Our client is revolutionizing real estate photo editing through AI. Currently processing over
20,000 images daily , we're on track to
scale to 100,000+ per day -while reducing delivery time from
30 minutes to under 10 . This exponential growth is powered by advanced machine learning, image processing, and real-time orchestration pipelines. We're looking for a senior backend engineer with deep experience in
Python ,
Kubernetes , and
high-performance inference systems
to lead this scale-up effort.
What You'll Do
You'll be responsible for scaling the backend systems that orchestrate a distributed inference pipeline built on Kubernetes, running 20+ CV/ML models per job with near real-time performance. You'll lead architectural decisions and implementation to ensure our systems are fast, robust, and production-grade. Scale and optimize our GPU-backed inference pipeline
in Kubernetes (GKE), improving throughput, reliability, and fault tolerance across tens of thousands of daily editing jobs. Build and refine Python services (FastAPI)
that orchestrate image workflows, manage job queues, and coordinate computer vision models. Design and implement new algorithms/models
for intelligent photo detection, sorting, and editing, and integrate them into our automated pipeline. Mentor and guide junior engineers , providing technical leadership on Kubernetes operations, backend structure, and ML system deployment best practices. Take ownership of deep performance work like
GPU memory tuning, pod autoscaling , and reducing cold-start latency across distributed workloads. What We're Looking For
6-10+ years
experience building and scaling complex backend systems, ideally involving real-time data or image processing. Expert-level Python proficiency , especially with FastAPI and production-grade backend architecture. Proven experience managing and scaling
Kubernetes clusters
(GKE or similar), ideally for machine learning or inference-heavy workloads. Familiarity with
model serving frameworks , such as TorchServe, TensorFlow Serving, or custom containerized inference flows. Background in high-throughput, performance-sensitive systems (e.g., real-time quality control, factory automation, media processing). High-agency, low-overhead
contributor: You identify problems, design solutions, and deliver them-quickly. Productivity mindset: You believe in shipping value daily or weekly, not monthly. Passion for performance, reliability, and engineering excellence-especially in the context of ML model execution and pipeline orchestration. Track record of mentoring and uplifting junior developers and engineers. Bonus Experience (Not Required)
Model training or tuning experience, object detection, style transfer, segmentation. GCP, Helm, Docker Compose, or GPU infrastructure tuning (NVIDIA MIG, CUDA/TF optimizations). Experience with asynchronous task queues (e.g., Celery, Redis queues) in production pipelines. Expectations & Work Culture
Location : In-office, Downtown Austin (Capital Factory). Time Commitment : 55-60+ hours/week, with presence required from
9am-6pm daily , and part-time effort on weekends. Our team moves fast-and we expect our senior engineers to lead that pace. Culture : This is a hands-on, high-output role. We're building category-defining technology with urgency. You bring the experience that helps us avoid unnecessary detours and scale with confidence. Why Join Us
Direct impact on a product transforming an entire industry. Opportunity to lead deep technical work on one of the most advanced image processing pipelines in real estate tech. Work with a mission-driven, fast-moving team obsessed with quality, automation, and user value.
Company Overview
Our client is revolutionizing real estate photo editing through AI. Currently processing over
20,000 images daily , we're on track to
scale to 100,000+ per day -while reducing delivery time from
30 minutes to under 10 . This exponential growth is powered by advanced machine learning, image processing, and real-time orchestration pipelines. We're looking for a senior backend engineer with deep experience in
Python ,
Kubernetes , and
high-performance inference systems
to lead this scale-up effort.
What You'll Do
You'll be responsible for scaling the backend systems that orchestrate a distributed inference pipeline built on Kubernetes, running 20+ CV/ML models per job with near real-time performance. You'll lead architectural decisions and implementation to ensure our systems are fast, robust, and production-grade. Scale and optimize our GPU-backed inference pipeline
in Kubernetes (GKE), improving throughput, reliability, and fault tolerance across tens of thousands of daily editing jobs. Build and refine Python services (FastAPI)
that orchestrate image workflows, manage job queues, and coordinate computer vision models. Design and implement new algorithms/models
for intelligent photo detection, sorting, and editing, and integrate them into our automated pipeline. Mentor and guide junior engineers , providing technical leadership on Kubernetes operations, backend structure, and ML system deployment best practices. Take ownership of deep performance work like
GPU memory tuning, pod autoscaling , and reducing cold-start latency across distributed workloads. What We're Looking For
6-10+ years
experience building and scaling complex backend systems, ideally involving real-time data or image processing. Expert-level Python proficiency , especially with FastAPI and production-grade backend architecture. Proven experience managing and scaling
Kubernetes clusters
(GKE or similar), ideally for machine learning or inference-heavy workloads. Familiarity with
model serving frameworks , such as TorchServe, TensorFlow Serving, or custom containerized inference flows. Background in high-throughput, performance-sensitive systems (e.g., real-time quality control, factory automation, media processing). High-agency, low-overhead
contributor: You identify problems, design solutions, and deliver them-quickly. Productivity mindset: You believe in shipping value daily or weekly, not monthly. Passion for performance, reliability, and engineering excellence-especially in the context of ML model execution and pipeline orchestration. Track record of mentoring and uplifting junior developers and engineers. Bonus Experience (Not Required)
Model training or tuning experience, object detection, style transfer, segmentation. GCP, Helm, Docker Compose, or GPU infrastructure tuning (NVIDIA MIG, CUDA/TF optimizations). Experience with asynchronous task queues (e.g., Celery, Redis queues) in production pipelines. Expectations & Work Culture
Location : In-office, Downtown Austin (Capital Factory). Time Commitment : 55-60+ hours/week, with presence required from
9am-6pm daily , and part-time effort on weekends. Our team moves fast-and we expect our senior engineers to lead that pace. Culture : This is a hands-on, high-output role. We're building category-defining technology with urgency. You bring the experience that helps us avoid unnecessary detours and scale with confidence. Why Join Us
Direct impact on a product transforming an entire industry. Opportunity to lead deep technical work on one of the most advanced image processing pipelines in real estate tech. Work with a mission-driven, fast-moving team obsessed with quality, automation, and user value.