Machine Learning Engineer

As a Machine Learning Engineer at Career.io, you will own ML products end to end, guiding work from problem definition through production deployment and the metrics used to confirm impact. The role is fully remote and focuses on data and entity resolution, retrieval, ranking and matching, plus applied LLMs and agentic workflows.

Responsibilities

• Own machine learning products end to end, taking initiatives from the initial problem framing to production delivery and the specific metric that demonstrates results.
• Own experiments including those that do not pan out, with responsibility for deciding when to stop rather than only launching initiatives.
• Build and maintain canonical datasets for titles, companies, skills, and industries, including content-addressed IDs, faceted taxonomies, and alias graphs accumulated across tens of millions of rows.
• Create rules-based resolution pipelines with LLM escalation, using the alias graph as a durable asset and aiming to reduce escalation volume over time.
• Run nightly agent loops that adjudicate ambiguous entities and propose structural changes, gated by invariant checks and blast-radius limits before committing updates.
• Handle job ingestion at scale across multi-source feeds, including deduplication, freshness, and the indexing economics needed to support retrieval.
• Build job matching v2 using a two-tower retrieval approach with cross-encoder reranking, trained on outcome labels rather than clicks, supported by hard-negative mining, propensity weighting, and impression-time logging.
• Learn mobility embeddings from observed career sequences to capture similarity when a text encoder cannot recover the relationship.
• Perform pivot feasibility analysis to determine where someone is, what moves are realistic, what is missing, and which intermediate roles have worked for peers.
• Fine-tune models where cost is justified against outcome labels, avoiding training for tasks a well-prompted frontier model can already handle.
• Develop agentic systems in production with human approval gates, producing reviewable artifacts for proposed changes and executing only after human sign-off.
• Implement continuous skills inference from work artifacts rather than relying solely on static documents.
• Create new product surfaces where correct behavior genuinely requires an LLM, and recognize when an LLM is not the right tool.
• Build evaluation infrastructure suitable for technical review, including time-forward splits, calibration, offline-to-online agreement, and careful handling of feedback-loop degeneration and survivorship bias.
• Design within real constraints including GDPR, EU AI Act high-risk classification for employment AI, and client data commitments treated as design inputs.

Requirements

• 5+ years of experience shipping ML systems into production, with the ability to name systems, define metrics before and after, and explain how you determined the model caused the change.
• Strong depth in both classical ML and deep learning (PyTorch or TensorFlow) applied to live products, not limited to notebooks and Kaggle-style datasets.
• Working fluency with LLMs in production, including retrieval, evaluation, prompt and context engineering, plus sound judgment to recognize when an LLM is not appropriate.
• Experience shipping with agentic coding tools such as Claude Code, Claude Design, or close equivalents, with the ability to reference what you built.
• Software engineering fundamentals strong enough to own deployments, including Python, Git, cloud experience (AWS), containers, and patience for messy human-authored self-reported data.

Technologies

PyTorch, TensorFlow, Claude Code, Claude Design, Python, Git, AWS, containers

Role Details

• Career.io is growing its machine learning team, and this role is part of how the organization builds products.
• A small product strategy team sets direction and priorities, while engineers own work end to end: discovery, design, build, ship, and the resulting impact.
• The role emphasizes autonomy within your domain along with accountability for outcomes.
• AI-native development is treated as the baseline, and engineers use Claude Code and Claude Design as default tools.
• This is a 100% remote/work-from-home role.

Nice-to-Haves

• Experience with entity resolution, record linkage, or taxonomy design at scale.
• Background in ranking, recommendation, or two-tower retrieval systems.
• Experience with sequence models on longitudinal or event-stream data.
• Exposure to embedding and vector retrieval systems in production.
• Experience with experiment design, causal inference, or off-policy evaluation.
• Experience with warehouse-native ML such as dbt, Snowflake, or similar tools.
• Domain experience in labor market, HR tech, or people-data.
• Open-source contributions or publications.