AI Resource Allocation Optimization Developer

Xivio Solutions

Remote · $60k – $80k / year · Full-time · Posted 5 months ago

This role is no longer accepting applications

Don't miss the next role like this

Create a free profile with your skills. We match you to new AI roles as they open, and companies hiring can contact you directly.

Create free profile, 2 minutes

Similar open roles you can apply to now:

See all open AI jobs

This role is no longer accepting applications

Don't miss the next role like this

Create a free profile with your skills. We match you to new AI roles as they open, and companies hiring can contact you directly.

Create free profile, 2 minutes

Similar open roles you can apply to now:

See all open AI jobs

About the role

Cloud compute is the biggest controllable cost for ML-heavy companies, and most of it is wasted on over-provisioned clusters, idle GPUs, and workloads running at the wrong time on the wrong hardware. We build the resource optimization layer that sits above Kubernetes and makes compute allocation intelligent. The core product is an allocation engine that predicts workload resource requirements, schedules jobs on the right instance types, and dynamically rightsizes running containers without disrupting the workloads. Our clients save an average of 31% on cloud compute costs, and the optimization decisions happen automatically without requiring engineers to read documentation about instance families. You'll own the demand forecasting models that power the allocation engine: predicting job duration and resource consumption from historical patterns, handling cold-start for new job types, and learning from allocation outcomes to improve future predictions. The models run inline with job submission, so latency matters — under 100ms to allocation decision.

Fully remote.

Responsibilities

Build demand forecasting models for ML job duration and resource consumption
Design cold-start strategies for new job types without historical data
Implement online learning loops that improve predictions from allocation outcomes
Optimize inference latency to under 100ms for inline allocation decisions
Build evaluation pipelines comparing predicted vs actual resource consumption
Collaborate with scheduling engineers on how predictions feed the allocator

Requirements

1+ year in ML engineering or software development with interest in infrastructure
Experience with resource optimization, scheduling, or capacity planning problems
Strong Python skills and familiarity with Kubernetes concepts
Understanding of online learning and feedback loop design
Comfort with latency-constrained ML serving