Control Risks is a specialist risk consultancy that helps to create secure, compliant and resilient organizations in an age of ever-changing risk. Working across disciplines, technologies and geographies, everything we do is based on our belief that taking risks is essential to our clients’ success.
Entrepreneurialism and risk-taking has long been baked into the core DNA of the firm, and over the last 50 years Control Risks has materially innovated in, and accordingly disrupted, several global sectors, including in the security, digital, and political risk domains.
Incubated by Control Risks, “Project Atlantic” is an AI Reg Tech start up that will fundamentally disrupt 25 years of compliance, due diligence, and commerce dogma, uniquely delivering confidence, connectivity and opportunity between global buyers, suppliers and other third parties, and in doing so “Accelerate the World’s progress to fairer, more transparent, and better business”.
Joining “Project Atlantic” means joining one of the most potentially powerfully disruptive AI startups to launch this year in the UK, backed by one of the most prestigious, credible and globally regarded firms in the sector. We plan on claiming the future, and need the best people in the industry to help us take it: are you one of them?

REQUIREMENTS

“Project Atlantic” is an AI start-up concept that in 2025 will come to life. As an AI/ML Architect, you will be one of the key early team members, responsible for leading the design and prototyping of our data-driven risk platform. You will own the technical architecture that powers:

• Knowledge graphs of entities and relationships
• Natural language processing pipelines for legal/news content
• A risk scoring engine that adapts as new data arrive

This is not a traditional data science role—we’re looking for a hands on builder and systems thinker who understands how to turn noisy data into actionable insight.

PROVEN EXPERIENCE WITH:

• Knowledge graphs or property graphs (Neo4j, RDF, or similar)
• NLP tools: spaCy, Hugging Face, Transformers, or LLM pipelines
• ML pipelines: feature engineering, model updating, retraining loops
• Risk scoring or rule-based decision systems (bonus if AML/FinCrime)
• Strong Python skills and comfort with cloud environments (AWS/GCP/Azure)
• Excellent communication skills and experience translating complex technical work into actionable product features
• A startup mindset—bias toward action, comfort with ambiguity, and passion for building from scratch

CONTROL RISKS IS COMMITTED TO A DIVERSE ENVIRONMENT AND IS PROUD TO BE AN EQUAL OPPORTUNITY EMPLOYER. ALL QUALIFIED APPLICANTS WILL RECEIVE CONSIDERATION FOR EMPLOYMENT WITHOUT REGARD TO RACE, COLOUR, RELIGION, GENDER, GENDER IDENTITY OR EXPRESSION, SEXUAL ORIENTATION, NATIONAL ORIGIN, GENETICS, DISABILITY, AGE OR VETERAN STATUS”

If you require any reasonable adjustments to be made in order to participate fully in the interview process, please let us know and we will be happy to accommodate your needs.

WHAT YOU’LL DO:

• Design and prototype the AI system architecture integrating:
• Graph databases (e.g., Neo4j) for entity linking and risk propagation
• NLP pipelines for processing structured and unstructured news, legal documents, corporate information, questionnaire information, and other reports
• ML-driven and rule-based risk scoring systems
• Build the data ingestion layer for structured and unstructured sources (e.g., news APIs, financial datasets, legal briefs)
• Define and implement early ontologies or schemas for risk entities (PEPs, suppliers, litigations etc)
• Design a scalable architecture that can evolve into a production-grade system
• Collaborate with product and business stakeholders to define MVP use cases
• Evaluate and recommend open-source tools, models, and infrastructure.

“Project Atlantic” is an AI start-up concept that in 2025 will come to life. As an AI/ML Architect, you will be one of the key early team members, responsible for leading the design and prototyping of our data-driven risk platform. You will own the technical architecture that powers:

• Knowledge graphs of entities and relationships
• Natural language processing pipelines for legal/news content
• A risk scoring engine that adapts as new data arriv