Use less nodes

In general, less nodes in the flow equals a faster process because each node adds overhead to the process.

For example, when possible, you should combine multiple Assign nodes. Every time IPA has to read a new node this results in CPU overhead on the server, and if you have logging turned on, then there is overhead to write out the variables from that node to the log. This consequently slows down the process. Another example is when you are using Infor Lawson IOS. You should use database relations to complete the entire query in one call instead of creating several queries.