> Markdown version of [/jobs/ext/3866652-data-scientist](https://www.wearedevelopers.com/jobs/ext/3866652-data-scientist). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Data Scientist

- **Company:** CADABRA LLC
- **Location:** United States (Remote available)
- **Experience:** Experienced
- **Contract:** Permanent contract
- **Skills:** Airflow, Information Engineering, Cursor, Python (Programming Language), Machine Learning, Natural Language Processing, Named Entity Recognition, Recommender Systems, SQL Databases, Unstructured Data, Cloud Platform System, GitHub Copilot, Large Language Models, Apache Spark, AI Coding Agents, Information Technology, Claude, Data Pipelines, Databricks
- **Published:** October 22, 2026
- **Apply:** https://arc.dev/remote-jobs/j/redirect/p8u4ssuo62

## About the Role

3-5 years' hands-on experience in data engineering, with strong Python and SQL.

Experience building and maintaining data pipelines, ideally with Databricks, Spark or a comparable cloud data platform.

Some exposure to - or a strong interest in developing - data science/ML techniques, e.g. classification, clustering, NER, embeddings.

Experience turning raw or messy data into structured, ready-to-use formats for downstream use.

Comfortable working independently and using AI coding tools to move quickly.

You are a great fit if you have:

Experience with personalisation, recommendation systems, or blending usage/behavioural data with content.

Exposure to using LLMs to extract structure (keywords, entities, topics) from unstructured text.

A degree in Computer Science, Data Science, Engineering or a related field - or equivalent experience.

Experience with orchestration tools such as Airflow, Dagster or Databricks Workflows.

## Description

They need a Data Scientist with a software or data engineering background who has moved into data science over the past few years and wants to keep developing in that direction. Or a generalist who's happy moving between engineering and data science work rather than a deep specialist in either. They'll soon be bringing in a lot of new data and new data types from external sources, which will make this data structuring and manipulation work even more central to the role.

Your day-to-day impact:

Turn raw, messy or unstructured data - article text, usage data, and new data arriving from external sources - into clean, ready-to-use formats for downstream products.

Apply light ML/NLP techniques (e.g. entity extraction, classification, tagging) to structure unstructured text.

Adapt existing pipelines to ingest and normalise new data types as they come online.

Bring in new datasets - scraped or paid-for/third-party - and merge them with our existing data to create new, value-add content sets.

Build personalisation logic that blends usage data with content suggestions, helping surface the right content to the right customer.

Work with product and editorial to define the signals and rules behind personalised recommendations and content creation.

Data Science tasks:

Use statistical and ML methods (e.g. classification, clustering, embeddings) to analyse usage and content data, surfacing patterns and insights that inform personalisation and product decisions.

Use LLMs to extract structure - keywords, entities, topics - from unstructured article text, and check the accuracy and quality of that output.

Prototype and test simple models or scoring logic (e.g. for recommendations or content matching), working with product to validate they add value before scaling.

Try different approaches to a data or personalisation problem, evaluate what works, and explain the trade-offs in plain terms to non-technical stakeholders.

Data Engineering tasks:

Maintain and extend our Databricks/Python pipelines - ingestion, transformation, scheduling and monitoring.

Identify new data sources relevant to the product and help scope their integration.

AI-Assisted Development:

Use AI coding tools (e.g. GitHub Copilot, Claude, Cursor) to build and iterate on data pipelines and structuring work quickly.

Work across product, editorial and commercial teams to understand data needs and communicate findings clearly to non-technical audiences.

Monitor data quality in production, flagging issues and improving tooling and documentation.