Applied MLFull-time
Machine Learning Engineer
Remote-first, Europe and Americas overlap
Turn research pipeline outputs into reliable ranking, retrieval, and evaluation systems that improve what omegaXiv runs next.
You will work on the models and evaluation loops behind ranking, semantic search, review quality, and automated research assistance, with a strong bias toward measurable product outcomes.
Apply or comment
Send a short intro, relevant work samples, and a concise explanation of the systems you have owned.
Apply for this roleShare feedback on GitHubWhat you will own
- Build and evaluate retrieval, ranking, and recommendation systems for problems, papers, and reviews.
- Improve model-assisted workflows for summarization, routing, and research quality checks.
- Own offline and online evaluation design, including benchmark curation and regression tracking.
- Collaborate with product and infra on serving constraints, latency budgets, and model rollout safety.
- Translate ambiguous research ideas into production experiments with explicit success criteria.
What we need
- Production experience shipping ML systems, not only training notebooks.
- Strong Python fundamentals and practical understanding of modern LLM and retrieval workflows.
- Experience with model evaluation, error analysis, and experiment design.
- Ability to balance product iteration speed with rigor around metrics and regressions.
- Comfort reading papers and extracting what is useful versus what is merely interesting.
First 90 days
- Baseline the current search and ranking quality with explicit evaluation slices.
- Ship one measurable improvement to retrieval or ranking quality in production.
- Define a durable evaluation harness for at least one ML-assisted product workflow.
Stack and environment
- Python
- LLM APIs and eval tooling
- Vector search
- Experiment tracking
- Product analytics
Nice to have
- Experience with semantic search, recommendation, or ranking models.
- Familiarity with LLM evals, tool use, or agent-style orchestration.
- Background in scientific domains, developer tools, or knowledge systems.
- Experience building internal platforms for ML iteration.
Working style
How we operate
We value engineers who can reason from first principles, keep systems understandable, and make tradeoffs visible. The team is small, so ownership is real and surface area is broad.
If your best work is at the intersection of product urgency and infrastructure rigor, you will likely fit well here.