Sorry but...uh, have you ever managed a production database with millions of users? Asking an AI to do something like normalize all the phone numbers would be completely trivial. Why don't people do it? Hmmmmm...
[edit] What might be nice would be if it wrote an execution plan about what exact queries it was going to run. That might be helpful because it could spot outliers. But when you're talking about updating millions of rows of data for something, you'd have to have a death wish. Even if you're just starting off a DB with some bunch of data you're getting from somewhere else, you'd want normalization you could trust.
If data cleanup needs to be deterministic, why is so much of it done by humans? Two humans also won't normalize the same table in the same way. The main thesis is that AI data cleanup is cheaper and less mistake-prone than the same work done by a bored intern.
noduerme•2h ago
Okay so I would find this useful if it would identify every weird case and then write a deterministic execution plan that covered all of them, that I could look at before running it. Like, show me every query in order I need to run to normalize some table. That itself would be bery useful. But I'd never let a bot loose on a production DB (at least an intern can only make mistakes slowly)
ZeljkoS•1h ago
Great!
If all plan queries are deterministic, the "Save recipe as Python" action saves the entire process as a deterministic script:
Honestly, you are probably not the target user, as you seem familiar with data cleanup scripts and AI agents. I think the target users would be:
A) Non-technical people in your org who clean data by hand because they don't know to automate it.
B) Technical people who want to test multiple queries and preview transformations before generating an AI-driven transformation flow (and are OK with it).
noduerme•3h ago
[edit] What might be nice would be if it wrote an execution plan about what exact queries it was going to run. That might be helpful because it could spot outliers. But when you're talking about updating millions of rows of data for something, you'd have to have a death wish. Even if you're just starting off a DB with some bunch of data you're getting from somewhere else, you'd want normalization you could trust.
ZeljkoS•3h ago
https://github.com/ZSvedic/TamedTable/blob/main/spec/test-ca...
And if an action can be done deterministically, then JS code will be generated:
https://github.com/ZSvedic/TamedTable/blob/main/spec/test-ca...
If data cleanup needs to be deterministic, why is so much of it done by humans? Two humans also won't normalize the same table in the same way. The main thesis is that AI data cleanup is cheaper and less mistake-prone than the same work done by a bored intern.
noduerme•2h ago
ZeljkoS•1h ago
If all plan queries are deterministic, the "Save recipe as Python" action saves the entire process as a deterministic script:
https://www.tamedtable.com/#save-py
Honestly, you are probably not the target user, as you seem familiar with data cleanup scripts and AI agents. I think the target users would be:
A) Non-technical people in your org who clean data by hand because they don't know to automate it.
B) Technical people who want to test multiple queries and preview transformations before generating an AI-driven transformation flow (and are OK with it).