AI-Powered Fraud Detection System Prototype
Budget / Salary$1,500–3,000
TypeFreelance project
LocationRemote
Posted2 hours ago
We are looking for an experienced **AI/ML developer** to build a prototype of an AI-powered fraud prevention platform for banks and financial institutions.
The system should detect multiple types of fraud, including:
* **Phishing and social engineering** where customers are tricked into giving away passwords or OTPs.
* **Account takeover** involving new devices, unusual logins and suspicious behaviour.
* **Online card fraud / card-not-present transactions.**
* **Unauthorized bank transfers** and unusual transaction patterns.
* **Suspicious beneficiaries and mule/ghost accounts.**
* Fraudulent transaction networks where funds move through multiple accounts.
The AI should analyse multiple signals together and produce a **fraud risk score (0–100)**, explain why an activity is suspicious, and generate an alert for the bank.
For example:
**New device → suspicious login → OTP requested → new beneficiary → unusual large transfer → suspicious recipient**
The system should recognise the combination of events and flag it as high risk.
### Prototype First
We specifically want to build a **working prototype first**, using simulated or publicly available data. It will not connect to real bank accounts at this stage.
The prototype should demonstrate how the AI could eventually integrate through APIs with a bank's existing systems without replacing the bank's existing app, core banking system or payment infrastructure.
The prototype should include a **simple fraud investigation dashboard** showing transactions, risk scores, alerts, suspicious accounts and the reasons for each alert.
Experience with **AI/ML, fraud detection, fintech, cybersecurity, transaction monitoring, anomaly detection and behavioural analytics** is highly preferred.
Please provide examples of similar projects, your proposed technology stack and your estimated timeline to complete the prototype.
The system should detect multiple types of fraud, including:
* **Phishing and social engineering** where customers are tricked into giving away passwords or OTPs.
* **Account takeover** involving new devices, unusual logins and suspicious behaviour.
* **Online card fraud / card-not-present transactions.**
* **Unauthorized bank transfers** and unusual transaction patterns.
* **Suspicious beneficiaries and mule/ghost accounts.**
* Fraudulent transaction networks where funds move through multiple accounts.
The AI should analyse multiple signals together and produce a **fraud risk score (0–100)**, explain why an activity is suspicious, and generate an alert for the bank.
For example:
**New device → suspicious login → OTP requested → new beneficiary → unusual large transfer → suspicious recipient**
The system should recognise the combination of events and flag it as high risk.
### Prototype First
We specifically want to build a **working prototype first**, using simulated or publicly available data. It will not connect to real bank accounts at this stage.
The prototype should demonstrate how the AI could eventually integrate through APIs with a bank's existing systems without replacing the bank's existing app, core banking system or payment infrastructure.
The prototype should include a **simple fraud investigation dashboard** showing transactions, risk scores, alerts, suspicious accounts and the reasons for each alert.
Experience with **AI/ML, fraud detection, fintech, cybersecurity, transaction monitoring, anomaly detection and behavioural analytics** is highly preferred.
Please provide examples of similar projects, your proposed technology stack and your estimated timeline to complete the prototype.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.