Machine Learning Engineer

Talent Inc.

WorldwideremotePosted 14 days ago

Skill Required

Machine-Learning-EngineeringSoftware-EngineerApplied-Machine-LearningAI-Product-DevelopmentMachine-Learning-EngineerMachine-Learning-Engineer-JobsAI-ML-EngineerMachine-Learning-Engineering-JobsApplied-Machine-Learning-EngineerLead-Machine-Learning-EngineerMachine Learning EngineerMachine Learningsoftware engineeringDeep LearningEngineeringTensorFlowSnowflakesimilar)buildingQuery OptimizationPyTorchPythondesignCloudGenerative AIdbtAWSGitandFirewallAIFulltime

Key highlights

Required experience: 5+ years shipping ML systems into production
100% remote / work-from-home role
AI-native development is the baseline; engineers ship with Claude Code and Claude Design as default tools
Engineers own products end to end with the autonomy of a founder and accountability for failed experiments as well as launches
Software stack: Python, Git, AWS, containers, PyTorch/TensorFlow; LLMs in production for retrieval, evals, and prompt/context engineering
Building inside constraints including GDPR and EU AI Act high-risk classification for employment AI

Role overview

Careerminds is hiring Machine Learning Engineers to own products end to end — from the problem, to production, to the metric that proves it worked. Engineers work with the autonomy of a founder inside their domain and the accountability that comes with it, including owning experiments that don't pan out and the call to kill them; the team prefers running four honest experiments and shipping the one that works over shipping four things that merely look fine on a dashboard. This is a 100% remote role where AI-native development is the baseline — engineers ship with Claude Code and Claude Design as their default tools, and the company wants people already working this way who can show the trail of repos, PRs, or shipped work built with these tools.

Responsibilities

Build canonical datasets for titles, companies, skills, and industries — the layer every application depends on, with content-addressed IDs, faceted taxonomies, and alias graphs accumulated across tens of millions of rows
Build rules-based resolution pipelines with LLM escalation, where the accumulated alias graph is the durable asset and escalation volume should fall over time
Run nightly agent loops that adjudicate ambiguous entities, propose structural changes, and get gated by invariant checks and blast-radius limits before anything commits
Build job ingestion at scale: multi-source feeds, deduplication, freshness, and the indexing economics underneath
Build Job matching v2: two-tower retrieval with cross-encoder reranking, trained on outcome labels rather than clicks, with hard-negative mining, propensity weighting, and impression-time logging
Develop mobility embeddings learned from observed career sequences — the similarity a text encoder can't recover, where roles like Claims Adjuster and Underwriting Assistant are substitutable despite sharing no vocabulary
Build pivot feasibility: given where someone is, what moves are realistic, what's missing, and which intermediate roles actually worked for peers
Fine-tune LLMs where it earns its cost — against outcome labels, not for tasks a well-prompted frontier model already handles
Build agentic systems in production with human approval gates: agents that analyze, propose changes as reviewable artifacts, and execute only after a human signs off — pushing further a pattern currently running against tens of millions of customer touchpoints a year
Build continuous skills inference from work artifacts rather than static documents — a problem several enterprise customers are currently solving for themselves, badly
Build new product surfaces where the right answer genuinely requires an LLM, with the discipline to notice when it doesn't
Build evaluation infrastructure you'd defend in a design review: time-forward splits, calibration, offline-to-online agreement, and honest handling of feedback-loop degeneration and survivorship bias
Build inside real constraints: GDPR, EU AI Act high-risk classification for employment AI, and client data commitments are design inputs here, not someone else's problem

Requirements

5+ years shipping ML systems into production — and you can name the system, the metric before and after, and how you knew the model caused the change
Depth in both classical ML and deep learning (PyTorch or TensorFlow) applied to live products, not notebooks and Kaggle sets
Working fluency with LLMs in production — retrieval, evals, prompt and context engineering, and the judgment to recognize when an LLM is the wrong tool
Already ships with agentic coding tools — Claude Code, Claude Design, or close equivalents — and can point to work built with them
Software engineering fundamentals strong enough to own your own deploys — Python, Git, cloud (the company runs AWS), and containers
Patience for genuinely messy, human-authored, self-reported data

Nice to have

Entity resolution, record linkage, or taxonomy design at scale
Ranking, recommendation, or two-tower retrieval systems
Sequence models on longitudinal or event-stream data
Embedding and vector retrieval systems in production
Experiment design, causal inference, or off-policy evaluation
Warehouse-native ML (dbt, Snowflake, or similar)
Labor market, HR tech, or people-data domain experience
Open-source contributions or publications

Benefits

100% remote / work-from-home role

Additional details

Careerminds is a leader in career transition and coaching solutions, helping organizations support employees through change while enabling workforce growth and development
Product portfolio includes market-leading Career Transition and Coaching Services as well as Progression, their application for Career Frameworks and progression planning
The machine learning team is growing
A small product strategy team sets direction and priorities; engineers own the work end to end — discovery, design, build, ship, and the result
Engineers have the autonomy of a founder inside their domain and the accountability that comes with it
Accountability includes the unglamorous half of ML — owning the experiment that doesn't pan out and the call to kill it, not just the launch
The team prefers running four honest experiments and shipping the one that works over shipping four things that all look fine on a dashboard
AI-native development is the baseline, not an aspiration — engineers ship with Claude Code and Claude Design as their default tools, and the leverage that creates is why one engineer can own a product end to end
The company wants people already working this way who want to push the ceiling higher, not people who are merely curious about AI
In the interview candidates will be asked to show the trail: repos, PRs, or shipped work built with agentic coding tools
Careerminds believes that diversity in thought and cultural background leads to better teams and stronger companies
Careerminds seeks talented, qualified employees regardless of race, color, sex/gender (including pregnancy, gender identity, and gender expression), national origin, religion, sexual orientation, disability, age, marital status, citizen status, veteran status, or any other protected classification under country or local law
Careerminds is an Equal Employment Opportunity Employer
Closing call to action: 'Come join our team. Together, we'll help others tell their career stories and land their dream jobs.'
Originally posted on Himalayas