Data Scientist - Fraud and Survey Optimisation
Job Description
Data Scientist - Fraud & Survey Optimisation
Location: Hybrid (2 days/week in London)
Rate: £500-£550/day (Outside IR35)
Length: 3-6 months
Start: Within 2 weeks (max 4-week notice period)
About the Role
We're working with a leading research and insights business that is tackling fraudulent data submissions across large-scale survey platforms. As part of a dedicated Fraud and Optimisation team, they're looking to bring in a contract Data Scientist to build and automate models that enhance the integrity and quality of survey data delivered to clients.
You'll be joining at a pivotal time, with the team focused on identifying response anomalies and developing scalable tools to filter out invalid data - ensuring higher-quality insights across their client base.
Key Fraud Challenges
You'll help detect and eliminate these common types of fraud:
-
Out-of-Country Fraud: Participants misreporting location to qualify for region-specific surveys.
-
Simulation: Individuals creating multiple profiles to access more surveys.
-
Status Inflation: Users falsely qualifying for more surveys by over-claiming attributes.
Your Contribution
-
Work with fraud analysts to understand patterns in historic and real-time data.
-
Build and automate anomaly detection models and a fraud scorecard to flag invalid responses.
-
Improve survey yield, efficiency, and data quality using statistical and machine learning techniques.
-
Attend whiteboarding and strategic sessions twice weekly (on-site in Reading or London).
Required Experience
-
Previous experience working in data science or advanced analytics roles in fast-moving or ambiguous environments.
-
Strong understanding of fraud detection, classification modelling, and data optimisation.
-
Experience working with financial, survey, or behavioural data (e.g. trading data, yield models, customer profiling).
Desired Skills and Experience