Staff ML Engineer

iFIT

$170k–$215k base + 15% bonus · Remote

The Problem

iFIT is building a personalization engine that feels effortless across every place a member discovers content—recommendations and search should work like a great coach: fast, relevant, and consistently improving. We’re looking for a Staff ML Engineer who can own the strategy and development of the recommendation system and search algorithms that power that experience end-to-end.

This is a rare seat where you’re not just “building models”—you’re shaping how personalization works across product surfaces, turning ambiguous goals into measurable outcomes, and partnering closely with Product, Engineering, Data, and Security to ship what matters.

The Work

• Own the recommendation + search personalization strategy across product surfaces—define the approach, prioritize experiments, and drive decisions from offline results to real user impact.
• Design and build end-to-end ML systems for ranking, retrieval, and personalization (from data pipelines and feature design through model training, evaluation, and serving).
• Improve relevance with a rigorous experimentation loop—set up experiments, interpret results, and iterate quickly when signals change (new content, new user behavior, new product surfaces).
• Partner with product and engineering to ship—translate model behavior into product outcomes, collaborate on integration, and ensure the system performs reliably in production.
• Collaborate with data and security on safe, scalable personalization—work through data access, privacy/security constraints, and operational requirements so the system can grow without surprises.

What You’ll Need

• Proven experience building recommendation and/or search systems that operate in production (not just research prototypes).
• Strong ML engineering fundamentals: ranking/retrieval concepts, feature engineering, model training/evaluation, and practical deployment considerations.
• Experience designing measurable improvements—you can connect model changes to user/product outcomes and run experiments that actually inform decisions.
• Strong software engineering skills: building reliable pipelines/services, writing maintainable code, and debugging complex systems.
• Clear communication: you can explain trade-offs and progress to product/engineering partners and align on next steps.
• Proficiency using Claude Code effectively to accelerate building and improving ML systems (e.g., code generation/refactoring, test automation, and debugging support).

Even Better If

• Experience with large-scale personalization (multiple surfaces, high traffic, and evolving catalogs).
• Familiarity with modern LLM-adjacent personalization workflows (e.g., using LLMs to enhance retrieval/ranking or building AI-native development workflows).

How Success Looks (First ~90 Days)

• You establish an end-to-end understanding of how personalization and search data flows, where key metrics are defined, and what the current system is optimizing for.
• You identify the highest-leverage opportunities for relevance improvements and propose an experimentation plan with clear decision criteria.
• You ship at least one meaningful production improvement (modeling, retrieval/ranking logic, feature/data quality, or evaluation), with results communicated to stakeholders.
• You put guardrails around reliability and data/privacy constraints so the system can iterate safely.

Who Thrives Here

• You enjoy owning the full arc—from strategy to shipped system—and you’re comfortable making calls when the path isn’t fully defined yet.
• You work best when you’re partnering closely with product and engineering to turn model ideas into real user impact.
• You bring rigor to experimentation and evaluation, but also move with urgency when the business needs answers.
• You collaborate across disciplines (data, security, product, engineering) without losing technical depth.

The Team

You’ll work with a cross-functional group focused on personalization across iFIT’s product surfaces. Expect tight collaboration with Product and Engineering on integration and user impact, plus close partnership with Data and Security on data access, privacy/security constraints, and operational readiness. The goal is simple: ship personalization that improves continuously and feels great to members.

Non-Traditional Backgrounds

If you’ve built and scaled recommendation/search systems in any environment—startup, enterprise, or open source—we’d still love to hear from you.