About the role
Responsibilities Build the semantic layer defining measures, dimensions and definitions the model may use Design and build the signal layer that detects client changes before bankers ask Own calculations and definitions the reasoning-layer LLM relies on Build agents and MCP servers behind the assistant, including tools and evals Build pipelines, services and data models from raw source data to served numbers Requirements Built production systems from scratch, writing code rather than configuring managed services Experience serving data to an LLM, including evals and retrieval issues Solid C#, Python or another backend language with tests Depth in SQL, Spark, Databricks/Delta Lake, Dagster/Airflow/Temporal or dbt Experience with CI/CD, Docker, cloud infrastructure and observability Preferred Qualifications Shaped data tools into a metrics layer, metric views or MCP server Set up CI/CD, Docker or observability for a team Benefits Equity offered Hybrid or fully remote work, with no return-to-office mandate About the Company Codat is an advisory intelligence platform for commercial banking, founded in 2017. It has powered over 350,000 connections to business financial systems.