GraphQL for Perfex CRM: Ask Exactly What You Need

The REST Over-Fetching Problem

If you’re building a frontend application on top of Perfex CRM data, you’ve encountered this pattern:

Request: “I need client name, email, and their latest invoice total.”

REST response: Returns 47 fields per client including internal IDs, timestamps, metadata, and columns you never asked for. Then you make a second request to get invoices. Then a third to correlate them.

Three API calls. 90% wasted data. This is over-fetching — and it’s why modern frontend teams demand GraphQL.

One Request. Exact Data. Zero Waste.

Data Builder adds a production-ready GraphQL endpoint to your Perfex CRM at /api/v1/graphql. You query exactly the fields you need, across multiple resources, in a single request:

{
  clients(filter: { active: 1 }, limit: 10) {
    userid
    company
    email
  }
}

Response: Only userid, company, and email. Nothing more. Your frontend gets exactly the data shape it expects.

Powerful Filtering Built In

Every table automatically generates filter input types with operator suffixes:

{
  invoices(filter: {
    status: 2,
    total_gte: 1000,
    total_lt: 5000,
    duedate_lt: "2026-05-01"
  }) {
    id
    total
    duedate
    status
  }
}

Supported operators per column:

• eq (equality, default)
• ne (not equal)
• gt, gte, lt, lte (comparisons)
• like (pattern matching)
• in (list matching)
• is_null (null check)

All filters combine with AND logic. Security-excluded columns (password, hash, token) are silently ignored in filters.

Full Mutation Support

GraphQL isn’t read-only. Data Builder supports create, update, and delete mutations with write scope enforcement:

mutation {
  createInvoice(input: {
    clientid: 42,
    total: 2500.00,
    status: 1
  }) {
    id
    total
    status
  }
}

Mutations are auto-generated from your table schemas. Input types are created automatically — no manual schema writing required.

Write scope enforcement: Mutations only appear in the schema when the API token has the corresponding permission. If a token has read-only access, mutation fields are invisible (not just blocked) — they don’t exist in the schema at all.

Production Safety Controls

GraphQL’s flexibility is powerful, but without guardrails it can be dangerous. Data Builder includes production-grade protections:

The admin API Explorer allows introspection for internal development — but external tokens see a locked-down schema surface.

Auto-Generated Schema

You don’t write GraphQL schemas manually. Data Builder scans your whitelisted tables and generates:

• Query types for every table
• Filter input types with all operator suffixes
• Mutation types (create/update/delete) with auto-detected primary keys
• Proper scalar mapping: MySQL BIGINT → GraphQL String, DECIMAL → String (financial precision preserved)

When you whitelist a new table, its GraphQL types appear immediately. No code generation step. No schema files to maintain.

Test in the API Explorer

The built-in API Explorer includes a GraphQL sandbox with:

• Sample queries for every table
• Live execution against your database
• Response preview with syntax highlighting
• PII randomization in sandbox mode — real query structure, randomized sensitive data

Frontend Should Query, Not Fight APIs

If you’re building a mobile app, a Vue/React dashboard, or a SaaS product on Perfex CRM data — GraphQL gives your frontend team the flexibility to move fast without backend bottlenecks.

One endpoint. Any data shape. Zero wasted bandwidth.

Frontend should query, not fight APIs.