Senior Machine Learning Engineer in London

Industry: NGO, Tech for Good

Confidentiality: NDA required before client details can be shared

About the opportunity

This is a rare opportunity. We’ve already placed several people into this organisation this year, and they love it. It’s a mission-driven not-for‑profit that works on real-world intelligence and research challenges with direct, tangible impact. In this role, what you build matters.

The team is growing fast. You’ll join a small but high‑calibre Data, Tools, and Technical Tradecraft function that runs like a miniature start‑up inside a serious analytical organisation. Engineers here have real creative freedom. You’re expected to own your problem space, propose new approaches and help shape how the organisation uses machine learning.

The core challenge is entity resolution at scale. The organisation’s Knowledge Base brings together a large corpus of structured and unstructured data. Much of it is incomplete, ambiguous and in multiple languages. Your job is to connect the people, organisations and other entities scattered across those sources into one accurate, trustworthy picture. The outputs feed directly into analytical work, so provenance and explainability matter as much as performance. It’s technically hard, socially meaningful and largely greenfield.

The role

You’ll report to the Data Lead. You’ll work closely with the Data Engineer and DevOps Engineer, and alongside researchers, linguists and other subject matter experts. You’ll own the entity resolution and de‑duplication strategy from first principles through to production.

This is a hands-on technical role. It’s ideal for an experienced engineer who loves the tech and wants to stay close to it rather than move into people management.

Key responsibilities

• Design, deploy and maintain the organisation’s entity resolution and de‑duplication strategy across the Knowledge Base
• Handle ambiguity from incomplete records, inconsistent naming and multilingual data
• Build production ML pipelines with a trust‑first approach, so every output has the provenance and accountability needed for analytical use
• Own the full model lifecycle, from problem framing and data preparation through evaluation, deployment and monitoring
• Turn non‑technical requirements from researchers and linguists into quantified performance metrics and evaluation criteria
• Work with domain experts to develop defensible matching rules alongside learned approaches
• Partner with the Data and DevOps Engineers on the pipelines and infrastructure behind your models

What we’re looking for

• Strong experience resolving entities across multiple datasets
• Experience with multilingual and/or multimodal datasets
• Production Python, with depth in ML libraries (PyTorch, JAX) and NLP ecosystems (spaCy, LangChain)
• Applying core NLP concepts to real‑world problems, not just research
• Confidence working directly with non‑technical stakeholders and subject matter experts
• Motivated by both technical challenge and real‑world social impact

Nice to have

• Experience with model deployment and MLOps tooling
• Cloud‑native deployment, ideally on GCP
• Experience with knowledge graphs or large‑scale record linkage
• Start‑up or scale‑up experience, owning ML problems end to end
• Experience working in or alongside blended analyst teams
• 5% employee / 6% employer pension contribution
• 23 days holiday + 8 bank holidays + Christmas shutdown (outside of holiday allowance)
• Enhanced maternity pay and occupational sick pay (4 weeks full pay)

#J-18808-Ljbffr