Arjun Patel

I cut my dev time roughly in half with AI pair programming — the workflow

Indie perspective: hosted, always. My time is the scarce resource, not the per-token cost.

Cache aggressively. Half of user questions repeat — we answer those from cache and the LLM bill drop...

Same observation. I route planning-heavy tasks to Claude and quick fix-it loops to GPT. Routing by t...