The 3 Vs of Data Innovation
Volume, velocity, variety — and the fourth V nobody listed. An Engatica briefing from the original interview library, updated for the LLM era.
The '3 Vs' framed a generation of data strategy talks in the original Engatica library. The frame still works; the contents of each V have been replaced.
The briefing
Volume is no longer about storing more — it is about deciding what is worth keeping as training and retrieval material, with provenance attached. Velocity now means freshness of context: an agent answering from last quarter's index is a liability with good grammar. Variety has expanded to the unstructured majority — transcripts, tickets, documents, calls — which LLMs finally made machine-readable at scale. The unlisted fourth V is veracity: in an era of generated content, knowing which of your data you can still trust is the whole game.
Questions we ask every guest on this theme
- Which of your data sources would you not defend in an audit?
- How fresh does your customer context need to be — honestly?
- What did LLMs make possible that your old stack could not do at any price?
See also the book library for the deeper technical shelf.