Case study

Integrations and performance for a Salesforce ecosystem

Client: a mid-sized US education company (anonymized) Role: Salesforce + Python developer Duration: 53 weeks

Apex REST/Bulk API Webhooks Starlette Django/DRF Postgres

Problem

Three connected issues:

1. A vendor integration between ChargeOver and Salesforce did not cover several required scenarios.
2. The checkout portal suffered timeout errors, traced to slow Salesforce triggers on synchronous HTTP calls from the front end.
3. The business needed to sync Salesforce data into Postgres without the rigid requirement that both schemas match.

What I did

• Built a ChargeOver-to-Salesforce sync handler using ChargeOver webhooks, Starlette, and Postgres, covering the scenarios the vendor integration missed.
• Built a generic Salesforce-to-Postgres sync that tolerates non-equivalent schemas.
• Helped build in-house Python Django/DRF backends for specialized apps (student portal and checkout portal).
• Fixed checkout timeouts by splitting the front-end HTTP calls into much smaller steps through the DRF backends.

Result

• Timeout errors eliminated on the checkout portal. They had almost blocked the checkout experience entirely before that.
• New ChargeOver scenarios supported without changing vendors.
• Reliable Salesforce-to-Postgres sync with flexible schema mapping — roughly 20,000 records synced incrementally every hour.