Glossary of terms

Term	Meaning
Client	A customer! but also sometimes used in context of a row in the client_config table, which includes: - Actual clients - Opportunities during data evaluation - Churned clients (we have an async process to remove them later) - Internal test accounts for various reasons
CCM / Client Config Manager	The system that is responsible for defining client dataset configs and building them. Originally it was just managing the configs but it has evolved into the step function process that builds client datasets and stores them in client_data in postgres. Typically a client config has sourcing criteria (see below) and field definitions that defines the fields in the dataset.
Data blocks	Elements of data that can be added to a client dataset for sourcing or fields in the dataset. There are 2 types of data block, static and dynamic. Static are those that dont need to be configured, eg are static for all clients. Basically these are built by DBT and boil down to a table with one row per company_id. Dynamic data blocks have per use / per client configuration that affects how the table of one row per company_id is assembled. the SQL code for dynamic data blocks is assembled depending on the config by commons/ccm/dynamicDatablocksand each defined configuration is added as a WITH into the dataset query. All datablocks used in a config are JOINED together based on company_id to assemble the main dataset.
Identity resolution	the process of resolving some identifiers for a company onto a 'real company' (or at least what we consider to be a company). E.g. a company may have many associated domains (some redirecting to others). Broadly there is no right answer here, however as much as possible we try to match our concept of a company to a single 'linkedin company profile'. This can be problematic as with larger companies they might have many linkedin profiles (eg amazon + AWS). Also its complicated by companies using non-primary domains in their profiles (eg. bitly links or careers sub domains).