Data Scientist

About Transak C-Corp

Transak is driven by the vision that any financial application should facilitate user onboarding anywhere globally with a single click. We power onboarding for financial apps through services like authentication, KYC, risk assessments, and fiat on/off-ramping. Our innovative infrastructure leverages blockchain and stablecoin technologies and is trusted by leading platforms such as MetaMask, Coinbase, Ledger, and Trust Wallet, supporting over 10 million users across more than 450 active applications. We have successfully secured over $37 million in funding from prestigious investors including Consensys, Tether, and Animoca Brands.

Role Overview

Your core mission as a Data Scientist at Transak will be to minimize fraud incidents while maximizing the approval rate for legitimate transactions. You will have autonomy in choosing the optimal methods—be it deterministic heuristics, machine learning models, AI agents, or any other innovative techniques—to tackle sophisticated fraudsters. As a senior member of a compact team, you will fully manage the end-to-end fraud detection and prevention processes. Your work will directly impact company revenue and user transaction experience by balancing fraud reduction and friction.

Challenges and Scale

• Access and analyze tens of millions of historical transactions, each with over 60 unique data fields.
• Utilize approximately 1,000 risk signals per decision sourced from 13 different providers.
• Handle transaction volumes reaching around 10,000 orders per hour during peaks, with transaction values pushing seven figures in those hours.
• Detect and respond to high-frequency, coordinated attack attempts involving tens of thousands of fraud attempts over days.
• Adapt models dynamically as adversaries evolve tactics, preventing static classification issues.
• Work with a largely undeveloped modeling environment that offers opportunities to implement state-of-the-art machine learning solutions.

Key Responsibilities

• Take full ownership of fraud, chargeback, and transaction risk modeling—from problem formulation and feature engineering to model deployment and threshold policy setting.
• Develop and optimize machine learning and signal-processing layers complementing external risk vendor solutions.
• Drive the strategic direction of risk modeling efforts and foster team alignment towards these goals.
• Serve cross-functional teams as a data scientist by enhancing data accessibility through tools such as the Snowflake data warehouse, self-service analytics platforms, dashboards, and internal data assistants.
• Address high-impact challenges related to product development, growth, and experimentation as they emerge.
• Advance internal AI adoption, including building coding assistants, developing internal copilots, and exploring innovative applications of large language models.

Candidate Profile

• Possesses strong foundations in mathematics and statistics, with enthusiasm for tackling complex quantitative problems.
• Proactive self-starter who independently identifies, scopes, and solves critical issues without needing direction.
• Competent engineer and analyst, proficient in Python and SQL, able to translate models from concept to production-ready systems.
• Demonstrates intellectual honesty regarding uncertainty and rigorously validates their work.
• Shows a genuine interest in cryptocurrencies, payment systems, fraud detection, and related data.

Preferred but Not Mandatory

• Notable academic achievements or competition successes such as strong university degrees, Kaggle rankings, or open-source contributions.
• Experience with fraud, risk management, payments, cryptocurrency, or adversarial machine learning contexts.
• Skills in deploying and monitoring production models, including managing model drift.
• Expertise in experimentation methods or causal inference techniques.
• Inclination to develop personal tools and automation solutions.

Why Join Us

• Work autonomously with a clear focus on delivering results rather than prescribed processes or fixed hours.
• Engage in an AI-first environment leveraging cutting-edge tools like Claude and Codex extensively in both engineering and data analysis.
• Be part of a small, high-trust, and technically sophisticated team that values rapid feedback and minimal bureaucracy.