Building agentic AI with advanced orchestration (multi-agent workflows, tool use, evaluation loops) that turns complex talent data into decisions.

• Training and fine-tuning our own models on unique data about jobs, skills and tasks, and benchmarking them against the best in our market.
• Joining an award-winning AI team (ACM RecSys Challenge 2026) that turns new ideas into innovative solutions with scientific rigour.
• Sharing our work and learning from industry peers through challenges, conferences and publications.

But it’s not all about creating high-quality products, we also very much value the company culture we have worked hard to create; built on trust, empathy & honesty ensuring our workforce is able to bring their full selves to work.

The Data Science team builds the AI models behind Beamery's Data Foundry, the common data and AI layer powering all Beamery products, and works hand in hand with the product teams to empower their features with AI (such as Talent Match and recommendations). At the heart of our work is the Beamery taxonomy, the language of work (skills, role titles and tasks) that every Beamery product speaks, built on more than a decade of research and kept current by the Taxonomist, our agentic system, so it reflects the real pulse of the job market. We combine LLMs, Transformers and embedding models with classic machine learning, always choosing the best balance between quality, latency and cost.

We see ourselves as pioneers in AI for HR Tech: in 2026 the team finished 1st in the industry track and 2nd overall out of 40 teams at the ACM RecSys Challenge.

You will join the team as a Senior Data Scientist and own our matching, recommendation and taxonomy mapping models, working closely with the product engineering teams that bring them to our customers. You will be responsible for the full cycle of your work, from defining how quality is measured and designing the models to running robust scientific experiments, tracking technical and business metrics, and writing and reviewing code. You will also advise the product teams on how to implement AI features with an evaluation-first mindset, and keep pushing the boundaries of Beamery's AI.

Key Responsibilities

• Build, improve and own our matching and recommendation models (e.g. candidate to job matching, internal mobility recommendations etc.).
• Apply an evaluation-first mindset: build golden datasets from real recruiter and user decisions, define quality metrics, design and run robust scientific experiments, track and report both technical and business metrics (using LLM-as-a-judge and human feedback), and communicate results and metrics clearly to the product teams.
• Write clean and tested Python code using AI coding agents (such as Claude Code or Codex), and review code with care, both from your peers and from AI agents, so the team's codebase stays consistent, readable and safe.
• Follow recent research (e.g. LLMs, embeddings, recommender systems, explainability, model evaluation) and bring the most relevant ideas into our product roadmap.
• Document your work and present it clearly to product, engineering and business stakeholders, and contribute to the team's research visibility (such as challenges, conferences and publications).

Required Skills and Experience

• A minimum of 2-4 years of experience as a Data Scientist, Senior Data Scientist (or a similar role) in an AI startup or an AI team, with a dynamic environment and a variety of use cases.
• Strong experience with recommender systems, search, ranking or matching (e.g. learning-to-rank, scoring models, Gradient Boosting Trees, embedding models).
• Fluency in evaluation: designing datasets and quality metrics (such as AUC, NDCG, precision and recall), and using LLM-as-a-judge and human feedback.
• A strong foundation in NLP and GenAI, including Transformers, LLMs and embedding models (training or fine-tuning them is a strong plus).
• High-quality software development skills in Python (clean design, testing, packaging, maintainable code etc.), practical knowledge of CI/CD, and hands-on experience with agentic coding, with the judgement to know when AI-generated code is right or wrong.
• The ability to simplify complex problems, to work with ambiguity, and to collaborate with data scientists, engineers, product managers and business stakeholders.
• A Master's degree or PhD, or equivalent experience, in Computer Science, Data Science, Mathematics or a related field.

• Experience with explainable or conversational recommendations.
• Experience in HR Tech, talent acquisition or workforce data.
• Publications, open-source contributions or strong results in ML challenges (e.g. RecSys, Kaggle).

• The chance to own core AI models that are used by some of the world's most forward-thinking organisations.
• A team that publishes, competes and is recognised in the research community, while shipping real products.
• Work with the latest AI technologies and tools, including AI coding agents as part of the daily workflow.
• Close collaboration with experienced data scientists, QA, platform and product engineers.

At this time, we are unable to provide Visa sponsorship for this role.

Beamery is for Everybody. Diversity and open expression are fundamental to us. We acknowledge the challenges in our industry and strive to develop an inclusive culture where everybody can contribute. We are dedicated to creating an inclusive environment for everyone, regardless of ethnicity, religion, color, sexual orientation, gender identity, race, national origin, age, disability status, or caregiver status. If, for whatever reason, you need us to make reasonable adjustments and adaptations to our recruitment process, please email

Visit our Diversity, Equality and Inclusion page to learn more about progress and commitments.