Graph Database Layer

How it works

The graph database layer is YInfra's relationship intelligence capability -- a layer on top of the relational data that learns and surfaces the patterns of connection that matter most to each vertical.

For a med spa, the graph models the relationship between clients and their treatment history, referral source, and referring client. Over time, it can surface which treatments drive the highest re-booking rates, which clients have referred the most new business, and what the typical path from first visit to high-value client looks like. This is not a report you run once -- it is a living map of the business's relational dynamics.

For MCA (merchant cash advance) and private lending, the graph models broker trees: which brokers introduce which borrowers, what each broker's historical close rate and deal quality look like, and which funders have participated in similar deals. The broker-borrower-investor triangle is a graph problem, not a table problem -- standard relational queries cannot traverse it efficiently. The graph layer makes a query like who else has this broker worked with an instant answer rather than a manual investigation.

For referral-based local businesses, the graph tracks referral networks: who referred whom, what the downstream value of each referring client is, and which marketing channels produce the highest-quality referral relationships.

The graph layer is built on Supabase's relational foundation with graph traversal patterns layered on top -- no separate graph database infrastructure is required for deployment.