Founding Causal AI Engineer — Remote, Equity

Referment; Greater London; Remote; GBP 90,000 - 150,000; Full time

Benefits offered by this job

Founding-team equity

Remote-friendly policy

Job summary

Referment is building an applied AI research lab and seeks a senior ML/AI scientist to own causal inference and experiments. You will work 70% on research, design, code and releases, with the rest leading domain efforts and shaping how methods reach the platform.

You'll join a small team, work directly with the founding team, and tackle open problems: causal modeling, evaluation design, and production-ready inference that informs strategy for client problems.

Qualifications

• 5+ years building AI/ML systems in production.
• Proven causal inference on business data (DiD, synthetic control, uplift, MMM, experiments).
• Strong statistics: uncertainty, calibration, experiment design.
• Experience designing benchmarks and trusted evaluation.
• Strong Python and SQL; AWS or Azure.
• Real LLM depth: RAG, tool-using, or text-to-SQL beyond demo; LangChain/LangGraph/LlamaIndex familiarity.
• Clear design docs before coding; explain cause and effect to CFO.
• Degree from a top-tier institution with AI-native learning.

Responsibilities

• Build the causal layer of the platform: map drivers, estimate effects, and refute estimates before use.
• Return effects as honest ranges with lag calibration.
• Design experiments with clients and translate forecasts into ranges.
• Own the attribution method that credits decisions with outcomes.
• Design evaluation: benchmarks, scoring rubrics and blind holdouts, plus trustworthy statistics.
• In first 90 days learn the platform and ship a causal estimate on a live client problem.

Skills

Python

SQL

Causal inference

Experiment design

Production ML

Education

Top-tier AI/ML degree

Tools

AWS

Azure

Databricks

LangChain

LlamaIndex

Job description

Referment is building an applied AI research lab and seeks a senior ML/AI scientist to own causal inference and experiments. You will work 70% on research, design, code and releases, with the rest leading domain efforts and shaping how methods reach the platform.

You'll join a small team, work directly with the founding team, and tackle open problems: causal modeling, evaluation design, and production-ready inference that informs strategy for client problems.