[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.]

[@will_the_ai_guy](/creator/twitter/will_the_ai_guy)
"I was paying $50/month for a fancy RAG DB setup for my docs I had a complex MCP server set up and everything then I realized u can just paste in the docs url and cursor will crawl it and create the docs index for u many such examples of entire SaaS products getting Jira ticketized by a larger company I guess this is just another lesson in how important it is to pick one tool master it and make sure ur spending a decent amount of time on "meta"' work - optimizing ur workflow and processes to make ur actual work more efficient"  
[X Link](https://x.com/will_the_ai_guy/status/1979688577874513983) [@will_the_ai_guy](/creator/x/will_the_ai_guy) 2025-10-18T23:18Z XXX followers, XX engagements


"Is any platform safe"  
[X Link](https://x.com/will_the_ai_guy/status/1979241811022647740) [@will_the_ai_guy](/creator/x/will_the_ai_guy) 2025-10-17T17:43Z XXX followers, XXX engagements


"Me: *Implements voice to voice chat pipeline with cloud providers for stt llm tss* Also me: I need to self host this whole pipeline to improve quality and latency *spends X days experimenting with locally hosted models* Me X days later: *goes back to initial implementation with cloud providers*"  
[X Link](https://x.com/will_the_ai_guy/status/1980118152479744224) [@will_the_ai_guy](/creator/x/will_the_ai_guy) 2025-10-20T03:45Z XXX followers, XXX engagements


"GCP salesmen rn"  
[X Link](https://x.com/will_the_ai_guy/status/1980283696260489717) [@will_the_ai_guy](/creator/x/will_the_ai_guy) 2025-10-20T14:43Z XXX followers, XXX engagements


"@vedantdotrpm apparently llm tokens"  
[X Link](https://x.com/will_the_ai_guy/status/1980287558031737228) [@will_the_ai_guy](/creator/x/will_the_ai_guy) 2025-10-20T14:58Z XXX followers, 1951 engagements


"DeepGram released a new flux model X weeks ago TLDR: Instead of separate modules for VAD(detecting when there is silence) Turn Detection (detecting if the silence if a pause or if the user is done speaking) and STT (transcribing what they say) they combine this all into X model and they even add triggers for speculative processing with my own benchmarks I'm already seeing around 500ms improvement using Flux vs my traditional pipeline bullish on conversational AI"  
[X Link](https://x.com/will_the_ai_guy/status/1980293427104559322) [@will_the_ai_guy](/creator/x/will_the_ai_guy) 2025-10-20T15:21Z XXX followers, XX engagements


"@dev_11217 i switch between gpt-5 gemini XXX pro and claude XXX sonnet"  
[X Link](https://x.com/will_the_ai_guy/status/1980302219326243078) [@will_the_ai_guy](/creator/x/will_the_ai_guy) 2025-10-20T15:56Z XXX followers, XX engagements


"@aarondotdev what about co-found deep tech startup in 20s"  
[X Link](https://x.com/will_the_ai_guy/status/1980414742775951404) [@will_the_ai_guy](/creator/x/will_the_ai_guy) 2025-10-20T23:24Z XXX followers, XX engagements