Search & Relevance Engineer →

NewsMesh; Northern (KY); Hybrid; USD 170,000 - 230,000; Full time

Benefits offered by this job

Remote across the US

Ownership of shipped work

Competitive compensation

Job summary

NewsMesh is building real-time, structured news data and an API used by developers. You will design and tune search relevance over a dynamic corpus, own evaluation metrics, and improve how entities and topics are extracted to keep filters reliable.

We value Python proficiency, experience with embeddings and ranking, and a rigorous, data-driven mindset. This is a remote US position with ownership and a small, senior team.

Qualifications

• Strong understanding of search and ranking.
• Rigorous approach to evaluation and metrics.
• Proficiency in Python and handling large datasets.

Responsibilities

• Design and tune search relevance over a large, constantly changing corpus.
• Own evaluation: define precision and recall, measure honestly, improve deliberately.
• Improve entity and topic extraction to keep filters trustworthy.

Skills

Information retrieval fundamentals

Evaluation discipline

Python proficiency

Embeddings / ranking / learning-to-ran

Deduplication / clustering / entity |r

Tools

Typesense

Elasticsearch

Vector DB

Job description

Our search runs over a corpus that changes every minute. Your job is to make it fast, make it relevant, and make the improvement measurable. That spans index design, query understanding, ranking, and the evaluation work that separates a real gain from a convincing demo.

About NewsMesh

NewsMesh turns the world's news into clean, structured, real-time data. We ingest thousands of sources continuously, enrich every article with category, topics, people, and country relevance, and serve it through an API used by developers and, increasingly, by AI agents that discover and pay for it on their own. We are a small, senior team, and the work is measured by one thing: whether the data is right, current, and easy to use.

What you'll do

• Design and tune search relevance over a large, constantly changing corpus
• Own evaluation: define precision and recall, measure honestly, improve deliberately
• Improve entity and topic extraction so filters stay trustworthy

What we look for

• Experience with production search (Typesense, Elasticsearch, a vector DB, or similar)
• Solid information-retrieval fundamentals
• Rigorous about evaluation, not just shipping a model
• Comfortable in Python and working with large datasets
• Experience with embeddings, ranking, or learning-to-rank
• Worked on dedup, clustering, or entity resolution

Compensation

Depending on experience, the expected pay range is $170K–$230K.

What we offer

• Competitive pay, in the range listed above
• Remote across the US
• Real ownership: you will ship work customers depend on
• A small, senior team with high standards and little process

How we hire

We read every application. If there is a fit, expect a short intro call, a practical exercise rooted in the kind of work you would actually do, and a few conversations with the people you would work with. A link to something you have built beats a resume.