Staff / Principal MLOps Engineer

Full-Time | Remote (US or Canada) or Contract (6 months, potential to convert)

Your Role

We're seeking a Staff / Principal MLOps Engineer to join our team. Our ML footprint has grown quickly alongside the business: batch models that process records in the backend, real-time models that serve recommendations to job seekers, and daily pipelines that process every available job across the US and Canada. The layer we have not yet built is the observability and traceability around all of it. Today, when a model regresses, a job fails, or a recommendation looks wrong, especially where LLMs are involved, tracing the cause and reproducing it takes far longer than it should. You will own that problem: assess our ML pipelines and data architecture with clear eyes, decide what to build and in what order, and then build it. This is a greenfield mandate, influencing production models and users. We are looking for a full-time hire, but are open to running this as a six-month contract, depending on fit and what you are looking for.

What You'll Own

• Assessment and plan: Evaluate our current pipelines, data architecture, and ML workflows, and produce a prioritized, opinionated plan for what needs to change and why.
• AI/ML observability: Architect our AI/ML observability and traceability from the ground up: model and data monitoring, regression detection, lineage, and the ability to reproduce a questionable recommendation on demand, including for LLM-based systems.
• Systems design: Design data and ML systems that are anchored in customer needs and built to last, with clear tradeoffs documented so the team can build on them.
• Implementation: Rebuild and harden pipelines, upgrade the data architecture, and ship the improvements.
• Reliability and standards: Raise the bar on reliability and data quality, establishing the patterns and practices the rest of the team can run with.
• Dependency and security hygiene: Keep the stack current and secure: framework and package upgrades across services and model images, and vulnerability remediation carried out without destabilizing production.

Required Experience

• Staff or principal-level experience in MLOps, ML platform, or ML infrastructure
• Experience standing up MLOps practice: CI/CD for models, experiment tracking, feature stores, and model monitoring
• Experience building AI/ML observability and traceability in production: detecting regressions, diagnosing failures, tracing a prediction back to the inputs that produced it, and reproducing issues after the fact
• Experience operating models in both batch and real-time serving contexts, with an understanding of how the reliability requirements differ
• A track record of walking into complex, fast-grown systems, diagnosing the real problems, and materially improving them
• Strong systems design ability: you can translate customer and product needs into durable, scalable architecture, write it down clearly, and stay close enough to the code to implement it yourself
• Depth in classical ML methods in production, plus practical exposure to LLM-based systems and what it takes to observe and evaluate them once they are live
• Deep experience building and operating production data pipelines and ML workflows at scale
• Fluency across the modern ML and cloud stack: orchestration, containerization, infrastructure as code, CI/CD, model serving, and monitoring

Bonus Points

• Experience evaluating and working with AI/ML observability or LLM evaluation vendors, including clear judgment on when to buy and when to build
• Background in mission-driven, workforce, or government-adjacent data environments
• Publications, presentations, blog posts, or other public artifacts showcasing your expertise and knowledge of best practices in MLOps
• Comfort mentoring and leveling up a small data and engineering team while you build

Our Tech Stack for Data

• Languages: SQL, Python
• Data orchestration and transformation: Airflow, dbt
• Data storage and warehousing: PostgreSQL, Redshift, MongoDB
• Machine learning and model serving: AWS SageMaker (PyTorch models, artifact upload to S3, model registration), serving real-time and batch inference
• Visualization and reporting: Looker, Quicksight
• Infrastructure: AWS (S3, Redshift), GitHub Actions for CI/CD

Your Education

Your alma mater isn't our focus. Your grit, hunger, and drive are. If you learn continuously, tackle challenges head-on, and know your strengths and gaps intimately, you're our person.

Location

[CA/US Remote] We are open to candidates living anywhere in Canada or the US. For candidates living in Toronto, our office is conveniently located at 325 Front St West (a short walk from Union Station). You are welcome to come in on a hybrid schedule.

Travel Expectations

Although this role is remote, you may be expected to travel up to once per quarter for off-sites and team gatherings.

Compensation

We are looking for a full-time hire as well as open to engaging this role as a six-month contract with potential to convert. For the full-time path, the base salary range is USD $170,000 to $215,000 for candidates based in the United States and CAD $170,000 to $220,000 for candidates based in Canada, regardless of location. As a remote-first company, we benchmark compensation to the national market for comparable roles at institutionally-funded startups, targeting the middle of the market. Bands are designed for the lifecycle of the role - where you enter the band reflects your applied experience and other criteria established by the hiring committee, with room to grow through the band as you grow in the role.

Hiring Journey

At FutureFit AI, our hiring process is designed to help you assess whether this role and our culture are the right fit based on your unique skills, mindset, and experiences. We move fast and work with intensity, so we want you to get a real sense of that from the start.

Each journey includes a mix of interviews and a performance challenge. For this role, that might look like:

1. Online Application
2. Initial Screen with Director of People & Culture
3. Interview with Hiring Manager
4. Performance Challenge
5. Final 1:1 Interviews
6. Final Decision

Generally, this entire process takes around 6 weeks, although the timing can vary due to specific candidate circumstances.

Ready to shape the future of work?

At FutureFit AI, we're not just building a company-we're transforming how talent and opportunity connect. Join our driven team united by a commitment to job seekers and the workforce ecosystems we serve.

Company Snapshot:

• Team: 30-50 across US and Canada (hubs in NYC and Toronto)
• Customers: Workforce development agencies and intermediaries, government agencies, employers
• Industry: SaaS/AI technology
• Funding: Bootstrapped 0-1, then raised funding led by JP Morgan
• Structure: Growth, Customer Success, Product, Engineering, Data, People & Culture, Finance & Operations

Our Core Principles

• Be Curious
• Drive to Outcomes
• Raise the Bar
• Speed Matters
• Own It
• We Over Me

Use of AI in Hiring

At FutureFit, we use artificial intelligence (AI) tools to make our hiring process more efficient, consistent, and equitable-never to replace human judgment. We use AI in the following ways:

• Screening support: AI may help us compare applications against the skills and experience required for a specific role. These skills are defined by the hiring team for each position. A human reviews each application, with the AI assessment as just one input.
• Interview support: In some interviews, we may use an AI notetaker to summarize the discussion so interviewers can focus on being present in the conversation.
• Insights, not decisions: AI provides data points to support our team's evaluation but does not make or recommend final hiring decisions. Every hiring decision is made by people.