Analytics Engineer / Data Scientist

We’re looking for an Analytics Engineer / Data Scientist to own our data transformation layer, analytics infrastructure, and emerging ML capabilities.

You’ll be responsible for how raw transaction data becomes business intelligence — from dbt models in Snowflake to automated reports, Power BI dashboards, and eventually predictive models that power real-time routing decisions.

You’ll work directly with the founding team, investigate payment performance issues hands-on, and build the analytical systems that merchants and internal teams rely on every day.

What You'll Own

Analytics Infrastructure (Day 1)

• dbt project with 50+ models across staging, facts, dimensions, and marts layers
• Snowflake data warehouse (DEV + PROD) processing millions of transactions across multiple merchants
• Incremental merge strategies, SCD Type 2 snapshots, and Jinja macro-driven code generation
• Data quality monitoring and anomaly detection models

Reporting & Product

• Working with Product to develop automated PDF reports for alert investigations, health checks, and BIN routing analysis
• Preparing data for consumption by Power BI dashboards
• Integrating our API to develop alerting for auth rate drops, decline spikes, and anomaly detection

ML & Predictive Analytics

• Decline prediction: identify which transactions are likely to fail before they are attempted
• Optimal MID routing: real-time scoring to determine the best processor based on BIN, amount, campaign, time of day, and card type
• Chargeback prediction: flag high-risk transactions before they become disputes
• Recovery optimization: predict which declined transactions are worth retrying and when
• Anomaly detection: automatically identify root causes when authorization rates drop (processor issues, campaign changes, fraud spikes, or new decline codes)

What We're Looking For

Must-Have

• Strong SQL: window functions, CTEs, aggregation patterns, null handling, and performance tuning. You should be comfortable writing queries that compute order-level deduplicated authorization rates across 10M+ transactions, grouped by MID, campaign, country, and card type.
• dbt experience: incremental models, snapshots, Jinja macros, testing, and documentation. You’ve built and maintained a production dbt project, not just completed tutorials.
• Python for analytics: Pandas, data manipulation, scripting, and report automation. This is analytics Python, not web development Python.
• Cloud data warehouse experience: Snowflake, BigQuery, or Redshift.
• 3-4 years credit card payments domain knowledge (with strong ability to learn quickly). You should understand—or be able to quickly learn—concepts like AVS, CVV, chargebacks, MID cascading, BIN routing, and authorization rates in card-not-present transactions.

Strong Preference

• Experience with ML or statistical modeling. You don’t need to be a PhD, but you should be comfortable with classification models (logistic regression, gradient boosting), feature engineering on transactional data, and model evaluation. Precision/recall tradeoffs matter heavily in payments.
• Power BI or similar BI tools, building dashboards, managing refresh cycles, and writing measures (DAX or equivalent).
• Experience in payments, fintech, or ecommerce analytics.