Credit Risk Data Engineer

kiwi; Santo Domingo; Remote; COP 291,281,000 - 420,739,000; Full time

Benefits offered by this job

Remote work

Job summary

Kiwi is seeking a Data Engineer to own the data layer behind our credit risk decisions and ensure data is available, clean, and timely for underwriting and renewal modeling. You will define data contracts, monitor data quality, and maintain end-to-end pipelines in Snowflake.

You will work with the data science and engineering teams in a fully remote setting across Argentina, Colombia, and the Dominican Republic, shaping the data architecture and governance for risk analytics.

Qualifications

• 4+ years of data engineering or analytics engineering with ownership of a production data platform.
• Expert SQL and strong data modeling skills, including dimensional models and point-in-time snapshots.
• Hands-on experience with a cloud data warehouse, ideally Snowflake.
• Experience with a transformation framework such as dbt, using version control, code review, and CI.

Responsibilities

• Own the inventory of every credit data source and define data requirements, grain, freshness, and landing location.
• Write data contracts with vendor integrations covering fields, types, null rules, and volumes.
• Monitor freshness, volume, fill rate, and schema for every source and model input, and set up alerts.
• Triage data incidents, identify root causes, and drive fixes to closure.
• Own the Credit Risk layer of Snowflake: raw, staging, mart, and feature tables; maintain data pipelines and transformations.

Skills

SQL

Data modeling

Python

Tools

Snowflake

dbt

Airflow

Python

Metabase

Job description

We are looking for a Data Engineer to own the data layer behind Kiwi's credit risk decisions. Your mission is to make sure Credit Risk always has the data it needs, when it needs it, in the right shape and in the right place.

You will own every data point that feeds an underwriting or renewal decision, from the moment a vendor returns it to the moment a model or analyst uses it. If an input breaks, goes stale, or drifts, you will be the first to know.

This is a new role. Today our data scientists check data quality and availability while also building models. We are creating this position so the data layer has a dedicated owner and issues are caught early, before they affect decisions.

Responsibilities

• Own the inventory of every credit data source, including TransUnion, Clarity, Plaid, Prism, and internal data from the app and loan servicing.
• Define, for each source, what data Credit Risk needs, at what grain, how fresh it must be, and where it lands.
• Write data contracts with the engineers who build the vendor integrations, covering expected fields, types, null rules, and volumes.
• Confirm that raw vendor responses are stored completely and can be reused for model development and audits.
• Monitor freshness, volume, fill rate, and schema for every source and every model input.
• Set up alerts for when a feature's fill rate drops, a vendor stops returning a field, or volumes deviate from baseline.
• Track input drift for production models alongside the data scientists.
• Triage data incidents: find the root cause, route the fix to the right owner, and track it to closure.
• Own the Credit Risk layer of our Snowflake warehouse, including raw, staging, mart, and feature tables.
• Build and maintain the pipelines and transformations behind application, loan, performance, and vendor data.
• Keep modeling datasets reproducible, so data scientists can rebuild a point-in-time training set without leakage.
• Maintain the tables behind the Credit Risk dashboards and governed metrics.
• Write automated tests for keys, duplicates, ranges, referential integrity, and reconciliation between sources.
• Maintain documentation and lineage: what each table and field means, where it comes from, and who uses it.
• Keep the owner registry up to date for every vendor, model, and model input.
• Support vendor oversight by checking SLAs and reconciling pull counts against vendor invoices.

Requirements

• 4+ years of experience in data engineering or analytics engineering, including time as the main owner of a production data platform.
• Expert SQL and strong data modeling skills, including dimensional models, slowly changing data, and point-in-time snapshots.
• Hands‑on experience with a cloud data warehouse, ideally Snowflake.
• Experience with a transformation framework such as dbt, using version control, code review, and CI.
• Python for pipelines, tests, and automation.
• A track record of building data quality tests and alerting, using dbt tests, Great Expectations, Monte Carlo, Elementary, or custom checks.
• Experience with an orchestration tool such as Airflow, Dagster, or Prefect.
• An ownership mindset: you notice problems before others do, follow them to the root cause, and close them out.
• Clear written communication: you can write a data contract, an incident note, or table documentation that others rely on.

Our technology

Our risk and data environment includes Snowflake, DBT, Airflow, Python, and Metabase, with credit bureau and bank data from providers such as TransUnion, Clarity, and Plaid. We value experience with the underlying data engineering challenges; prior use of every tool in this stack is not required.

Nice to have

• Experience in fintech or lending, especially with credit bureau data (TransUnion, Experian, Clarity) or bank data (Plaid).
• Experience supporting data scientists with feature tables and training datasets, including avoiding leakage.
• Familiarity with model monitoring concepts such as PSI, drift, and fill rates.
• Exposure to lending compliance, such as FCRA, adverse action, and data retention.
• Experience with BI tools such as Metabase.
• Bilingual in Spanish and English.

What we offer

• The opportunity to work on critical financial products with direct impact on customers and business growth.
• Full ownership of the Credit Risk data layer and the opportunity to shape how it evolves.
• Meaningful challenges across data pipelines, data quality, monitoring, and modeling datasets.
• An environment where AI is becoming a core part of how we work and build.
• A collaborative multidisciplinary team across Credit Risk, Data Science, Engineering, and Product.
• 100% remote — [Argentina, Colombia, Dominican Republic].