Remote Full Stack Engineer: ML Orchestration Platform

Onward Search; New York (NY); Remote; USD 83,000 - 165,000; Full time

Benefits offered by this job

Medical, Dental, and Vision Insurance.

Life Insurance.

401(k) Program.

Commuter Benefit.

eLearning and Ongoing Training.

Education Reimbursement.

Job summary

Onward Search is seeking a Full Stack Engineer for a remote, in-region contract to build and operate a high-scale ML orchestration platform. You will enable online inference, model evaluation, and access to model-serving infrastructure.

You will work with product teams to connect models with features and serving infrastructure, ensuring low-latency, highly available services on Kubernetes, while applying functional programming patterns in Scala or similar languages.

Qualifications

• Back-end engineering experience working on machine learning systems.
• 5+ years building and operating distributed, low-latency production services.
• Expertise with Scala or another functional programming language.
• Understanding of production ML inference: model serving, feature retrieval, batching, latency/throughput trade-offs.
• Hands-on experience operating services on Kubernetes, including deployments, scaling, resource management, and production debugging.

Responsibilities

• Build and operate a high-scale ML orchestration platform.
• Support online inference and model evaluation workflows.
• Connect product teams with models, features, and serving infrastructure.
• Maintain reliable, highly available services for buyer/seller experiences.
• Collaborate with internal users and engineering partners to improve the platform.

Skills

ML systems

Distributed services

Scala or FP language

Kubernetes

Performance/latency

Job description

Onward Search is seeking a Full Stack Engineer for a remote, in-region contract to build and operate a high-scale ML orchestration platform. You will enable online inference, model evaluation, and access to model-serving infrastructure.

You will work with product teams to connect models with features and serving infrastructure, ensuring low-latency, highly available services on Kubernetes, while applying functional programming patterns in Scala or similar languages.