So company gave many of us claude code. And they are stupid as a rock, no restriction on what or how we use it. I think they have no idea about what it can actually do.
Can I abuse this somehow. Should I run trading with it? Should I make it reed some poision every day on a loop?
Any ideas?
On paper I am a data analyst. In reality I am just a janitor.
Background: Very strong python with pyspark and polars, also some actual programming like library creation. Mediocre SQL. Databricks and local developement. Used power BI but gave up, rather create a dash or shiny app than power bi or Tableau or databricks dashboard.
Issue: company data is flaming garbage, not huge but wide, very wide. Company is manufacturing, market leader, but had no data anything until last year. My job is "delivering insights" but reality is I am shoveling garbage into nicer piles of shiny garbage. I spend days after day just figuring out how to join some table and what basic filter to apply. So asking 8 people and trying it myself. All my reported KPIs have sidenotes for data quality issues that I know of.
Agentic is no help, how would Claude know the solution for a data model if no living man at the company does. Garbage in garbage out.
When I joined they even asked for some ML, you can guess how much of that I touched.
I am getting tired, boss. I should switch right? Anybody in the same boot?