Databases · MySQL
Connect MySQL. Use business data where it is needed.
Analyse CRM records, provide stock data or exchange records with a business application. MySQL can serve as a source or target for onflAI. Tables, keys and write permissions define the actual process.
A CONCRETE EXAMPLE
What your requirement could look like.
“Transfer HubSpot companies into MySQL. Update existing records by HubSpot ID and include only the agreed fields.”
A customer dataset ready for analysis, without creating extra copies of the same company on each run.
Planning example, not a customer case study. Execution, scheduling and scope depend on connectivity, approval and plan.
01
Prepare access
- Identify the MySQL version, host, port, database and required tables or views. Verify reachability from the execution environment.
- Use a dedicated account with minimal read or write permissions. Configure TLS and certificate checks to match the server and driver.
- Store credentials separately. A schema and anonymised samples are enough for planning; no full production dump is needed.
02
Define fields and rules
- Map external IDs to unique keys. Decide whether records should be inserted or existing ones updated.
- Check decimal values, character encoding, dates and NULL rules. Do not confuse database-generated IDs with source identifiers.
03
Check before production
- Test duplicates, missing required values and interrupted transfers. Related changes need an agreed transaction boundary.
- Define change indicators and reasonable read volumes for recurring runs. Review query load and indexes with IT before a large import.
What does a MySQL connection look like in practice?
The product tour shows HubSpot fields mapped to MySQL as a concrete example. Other cloud services or custom web services can also serve as the source. Database version, network access and permissions are aligned with your connection.
Where onflAI helps
You describe the outcome and select the relevant data. AI assists with configuration, mapping and validation. Credentials remain separate and are processed in encrypted form by the platform. After review and human approval, IOnbrix executes the data process, not AI.
A validated configuration does not replace an end-to-end test with your systems. Verify network access, the correct business outcome and error behaviour before production.