GlossaryKind of term: Manufacturing termIntegration
What is an integration between systems?
An integration is a configured, automated data exchange between two systems, through an API, a file exchange or middleware.
Nobody has to retype the data: an item is created in the ERP as soon as a quote is accepted, hours from an MES arrive in post-calculation on their own. Technically it runs through an API, through files placed and read at set times, or through middleware connecting several systems. The exchange can run one way or both ways.
What does an integration depend on?
An integration only works when both systems mean the same thing. That takes agreements on which system governs which data, how fields map onto each other (item number, unit, revision, currency), what happens on an error or a duplicate record, and how changes on either side are tracked. An update to either package can break an integration, so maintenance is part of it.
How does an integration work in practice?
Double entry costs estimating and the sales office time and leaves room for differences between systems. An integration can remove that, but needs configuration, coordination with the vendors of both packages, and maintenance. Many companies therefore start with a manual export and import of files, and automate once the flow is stable and large enough.
What is an integration often confused with?
An integration is not a manual export: with an export someone moves a file across, with an integration it happens by configuration, without manual work, even when a file sits in between. Nor is an integration a digital thread: an integration joins two systems, a digital thread is the continuous chain of data across a part's life.
How does data get from Blake into your other systems?
Blake issues a quote version as a PDF, with CSV and JSON export for your ERP or spreadsheet. Machines and their hourly rates, materials and your own purchase prices are held in Blake itself.
Frequently asked questions
What is the difference between an API integration and a file-based integration?
With an API the systems exchange data directly, often per record and shortly after a change. A file-based integration hands over a file at set times. An API is more direct; files are simpler to set up and to check.
What should you agree before connecting two systems?
Which system governs which data, how fields and units correspond, how revisions and item numbers are matched, and what happens on an error. Without those agreements an integration swaps retyping for sorting out differences.