Internship

Data Science Intern

DI

Posted on 4/16/2025

Socure

Socure

201-500 employees

Digital identity verification and fraud prevention

No salary listed

Remote in USA

Category
Data Science
Data & Analytics
Required Skills
Scikit-learn
Python
Data Science
Tensorflow
Pytorch
Pandas
NumPy
Requirements
  • Currently pursuing a Bachelor’s, Master’s, or Ph.D. in Data Science, Computer Science, Statistics, Mathematics, or a related field.
  • Experience working with large datasets and knowledge of data cleaning, feature engineering, and model evaluation techniques.
  • Strong Python programming skills; familiarity with Pandas, NumPy, PySpark, or other scalable data processing tools.
  • Familiarity with ML frameworks such as scikit-learn, TensorFlow, or PyTorch.
  • Strong problem-solving skills and the ability to work independently or collaboratively.
  • Effective communication skills (both written and verbal).
Responsibilities
  • Work alongside data scientists and engineers to develop and optimize methods for fraud detection and device recognition.
  • Analyze large datasets to extract meaningful insights, detect patterns, and drive data-driven decision-making.
  • Develop and evaluate features derived from browser, app, and IP-based activity for device recognition and fraud detection.
  • Assist in the research and implementation of new machine learning algorithms and techniques.
  • Collaborate with cross-functional teams, including product, engineering, and analytics, to support business objectives.
  • Develop data visualizations and reports to communicate findings to both technical and non-technical stakeholders.
Desired Qualifications
  • Experience in fraud detection, anomaly detection, or adversarial machine learning.
  • Knowledge of ML techniques for high-cardinality features (e.g., frequency encoding, embeddings, target encoding).
  • Understanding of IP address behavior, device/browser data, or network-based feature engineering.

Socure focuses on digital identity verification and fraud prevention for financial institutions like banks and credit unions. It uses machine learning and advanced algorithms to analyze customer data, helping organizations quickly verify identities and reduce fraud risk. One of its main products, the Sigma Fraud Score, detects fraud by comparing transactions against a database of known fraudulent users. Socure operates on a subscription model, providing clients with continuous access to its tools while ensuring compliance with regulatory requirements.

Company Size

201-500

Company Stage

Debt Financing

Total Funding

$744.1M

Headquarters

New York City, New York

Founded

2012

Simplify Jobs

Simplify's Take

What believers are saying

  • Socure's subscription model ensures a steady revenue stream and continuous service updates.
  • Strategic executive hires are set to drive innovation and market alignment.
  • The launch of AI Copilot offers a competitive edge in digital identity verification.

What critics are saying

  • AI-powered scams using deepfakes challenge Socure's digital verification systems.
  • First-party fraud remains a significant challenge despite new risk scores.
  • Evolving BEC attacks require continuous adaptation of fraud prevention strategies.

What makes Socure unique

  • Socure's Sigma Fraud Score is renowned for its fraud detection accuracy.
  • The Identity Manipulation Risk Score is a pioneering tool against first-party fraud.
  • Socure's AI Copilot enhances global watchlist screening and monitoring capabilities.

Help us improve and share your feedback! Did you find this helpful?

Benefits

Competitive salary & equity

Comprehensive medical, dental, & vision coverage

FSA

Parental leave

401k

Life insurance

PTO

Growth & Insights and Company News

Headcount

6 month growth

4%

1 year growth

0%

2 year growth

0%
PYMNTS
Mar 6th, 2025
Socure Introduces New Risk Score To Combat Repeat First-Party Fraud

Socure has introduced a new tool designed to combat repeat first-party fraud. The new Identity Manipulation Risk Score, which is now embedded within Sigma First-Party Fraud, uses artificial intelligence (AI) and real-time data from Socure’s fraud intelligence network to help organizations detect bad actors at account opening and throughout the customer lifecycle, the company said in a Thursday (March 6) press release. “Identity manipulation is at the core of predicting first-party fraud, as most fraud prevention solutions fail to detect it,” Ori Snir, head of product management, fraud and identity solutions at Socure, said in the release

FF News
Mar 6th, 2025
Socure Introduces First-Ever Identity Manipulation Risk Score For First-Party Fraud

Socure, the leading AI-driven platform for digital identity verification, fraud prevention, and sanctions screening, today announced the launch of its Identity Manipulation Risk Score, the first-ever cross-industry predictive risk score designed to stop repeat first-party fraud abusers from exploiting the digital economy at scale.This AI-powered capability is embedded within Sigma First-Party Fraud, Socure’s innovative solution that leverages the largest cross-industry first-party fraud consortium to perform real-time analysis of dispute histories, payment denials, and account closures across millions of identities and billions of transactions. Socure’s first-party fraud consortium spans major financial institutions, fintechs, payment platforms, sports betting companies and merchants providing an unmatched network of fraud intelligence from diverse sectors. Sigma First-Party Fraud provides organizations with critical risk signals, real-time risk monitoring and actionable intelligence to detect bad-faith actors at account opening and throughout the entire customer lifecycle.A Breakthrough in First-Party Fraud PreventionFirst-party fraud, or the use of one’s own identity to open an account and commit a dishonest act for personal or financial gain, is costing businesses over $100 billion annually. Unlike third-party fraud, which involves stolen or synthetic identities, first-party fraudsters exploit their real, verified credentials, making it far more difficult to detect. What’s worse, in a recent survey Socure found that nearly half (49%) of those who committed first-party fraud in 2024 said they did so because they got away with it in 2023, proving that repeat abuse is a growing problem without a deterrent.With the introduction of the Identity Manipulation Risk Score, organizations can now leverage Socure’s Sigma First-Party Fraud solution to quantify the likelihood of identity manipulation at critical decision points, including account opening, high-risk transactions, and dispute resolution. Identity manipulation behaviors include consumers applying for services with different emails or phone numbers to exploit multiple organizations using their own true identity

FF News
Mar 6th, 2025
Socure Introduces First-Ever Identity Manipulation Risk Score for First-Party Fraud

Socure, the leading AI-driven platform for digital identity verification, fraud prevention, and sanctions screening, today announced the launch of its Identity Manipulation Risk Score, the first-ever cross-industry predictive risk score designed to stop repeat first-party fraud abusers from exploiting the digital economy at scale.

Neuron
Mar 3rd, 2025
Socure Launches AI-Powered Assistant for Global Watchlist Screening

Socure has launched an innovative AI assistant that enhances its Global Watchlist Screening and Monitoring solution, revolutionizing digital identity verification.

PYMNTS
Mar 2nd, 2025
Scammers Tap Social Media Selfies And Genai To Fuel Fraud

The wide availability of selfies on social media has helped fuel artificial intelligence (AI)-powered scams. Speaking to Bloomberg News on Friday (Feb. 28), Center for Strategic International Studies researcher Julia Dickson said cybercriminals are stealing profile pictures and personally identifiable information or purchasing it from underground marketplaces. From there, scammers use the data to craft deepfakes that let them get around digital verification systems and know-your-customer (KYC) measures