RAG got better when I stopped treating every question the same.
Extractors for the exact stuff first. Retrieval + LLM second.
Fuzzy is easy. Exact is the bar.
What do you route deterministically?
I've launched a slack channel for AI enthusiasts, developers, vibe coders, vibe operators, and anyone who is a little bit sick of the anti-AI sentiments across the other popular slack and discords
https://t.co/A4uAmsQPwI
If you want to help build a community around AI lets do it together
🧠 The real bottleneck in AI isn't the model.
It's the data.
Manual labeling = slow, brittle, and tough to scale.
We need to treat data like code:
✅ Versioned
✅ Reusable
✅ Automated
👇 How are you solving this?
#DataCentricAI#MLOps#TrainingData#MLDev#AIEngineering
CA just became the 4th largest economy in the WORLD.
In 2023, we contributed $83 billion to the fed govt.
Texas? Took $71 billion.
71% of America's economic output comes from dem areas-- places with longer lifespans, higher wages, & lower murder rates.