Principal AI Infrastructure Engineer

HubSpot; United States; Hybrid; USD 286,000 - 457,000; Full time

Job summary

HubSpot is seeking a Principal Machine Learning Engineer to help shape the base AI infrastructure powering Breeze Engine and Aviator, HubSpot’s internal frameworks for building and running agents. This is a high-scope, infrastructure-focused role that creates leverage across the company.

You’ll design and build agent runtime, evaluation systems, quality signals, model routing, and tuning workflows, collaborating with infrastructure, product, backend, and ML teams.

Qualifications

• Senior technical leader with deep experience in production ML, LLMs, or AI infrastructure.
• Experience building systems at scale where reliability, latency, cost, and quality matter.
• Ability to think in platforms and influence across infra, product, backend, and ML teams.

Responsibilities

• Build infrastructure for agent runtime, evaluation, quality measurement, and model optimization.
• Create tooling to show eval strengths, coverage gaps, and unanswered user questions.
• Design signal pipelines to surface user frustration and quality issues before CSAT drops.
• Help define an HubSpot-specific AI benchmark for evaluating model performance on real workloads.
• Build systems to evaluate new models across agents to improve quality, reduce cost, and support routing.
• Make fine-tuning and optimization repeatable for HubSpot’s AI use cases.

Skills

Production ML

LLM

AI infrastructure

Platform thinking

Cross-team influence

Tools

Eval frameworks

Model serving

Fine-tuning

Signal extraction

Experimentation

Feedback loops

Model optimization

Job description

HubSpot is seeking a Principal Machine Learning Engineer to help shape the base AI infrastructure powering Breeze Engine and Aviator, HubSpot’s internal frameworks for building and running agents. This is a high-scope, infrastructure-focused role that creates leverage across the company.

You’ll design and build agent runtime, evaluation systems, quality signals, model routing, and tuning workflows, collaborating with infrastructure, product, backend, and ML teams.