We inherited a system with a few hundred tables, no documentation and nobody left who had built it. Working out how anything joined to anything else was going to be weeks of guesswork. Peliqan has a template app that detects foreign key relations for you, so we ran that first and got a usable picture of the relational model without reverse engineering it by hand. From there we worked through the Grid view adding descriptions to the tables and columns that mattered, so the next person would not have to repeat the exercise.

A client's application needed a clean feed of data that lived across three of their systems. Rather than build a service to sit in front of all that, we prepared the dataset with SQL and published it as an API endpoint straight from Peliqan. Their developers consumed it directly.<br><br>That took the whole middle layer out of the project. No server for us to run, no auth to write ourselves, and when they asked for two extra fields we changed the query and it was live

This is the feature we did not know we needed. Peliqan builds the lineage for your tables and queries on its own, so you can open a table, go to Details, and see both where the data came from and what is sitting downstream of it.<br><br>We found out the hard way. Someone was about to drop a column from a query that half a dozen things depended on, and the lineage view showed exactly what would have broken before we ran it. Previously that was tribal knowledge, and the answer was usually to change it and see who complains.<br><br>You can also annotate your own Python scripts so they show up as nodes in the graph, which means the custom steps appear rather than leaving gaps in the chain.