API‑triggered Process Model using Write to Data Store (MySQL) takes 90+ seconds for ~25 inserts into a single table

Hello Community,

I am facing a performance issue with an API‑triggered Process Model that writes data to a MySQL database using the Write to Data Store Entity smart service.

 Scenario

• An external system invokes an Appian Web API.
• The API starts a Process Model.
• The Process Model creates a case by inserting ~25+ rows into a single database table.
• Inserts are done using Write to Data Store Entity (CDT‑based).
• The API response is sent only after the Process Model completes.

 Issue

• Total API response time is 90+ seconds.
• The process completes successfully and returns "Successful", but the latency is too high for synchronous API calls.
• Since all records are inserted into one table, this delay seems unexpected.

 Current Design 

• API → Process Model
• Multiple Write to Data Store Entity executions (likely one per record / loop)
• Inserts are mostly sequential
• Backend database: MySQL

 Questions

1. Is it expected for Write to Data Store Entity to take this long when inserting ~25 rows into a single table?
2. Does Appian open a separate transaction / commit for each Write to Data Store execution?
3. Would performance improve by:
   • Passing all 25 records at once in a single Write to Data Store node?
   • Using a database stored procedure for bulk inserts?
4. In API scenarios, is it recommended to:
   • Immediately respond to the API
   • And continue DB inserts asynchronously in the background?