... Vibe coding is about moving fast from idea to production without getting stuck on infrastructure. A simple stack that works well: Frontend and Backend: @Google Antigravity , @Cursor , @Claude Code Databases and RAG: @Pinecone Cloud and DevOps: Bult.ai LLMs: @OpenAI , @Google Gemini, @Hugging Face , @Ollama Bult.ai handles deployment, infrastructure, SSL, domains, scaling, and CI CD, so you can focus on building the product, not managing servers. Curious what stacks other vibe coders are using ...

... organization - Check for Updates Feature: Manual update checking from the application menu - App Version Display: Version number displayed at the bottom of settings page - Platform Information Caching: Performance improvements for platform detection and telemetry Changed - Infrastructure Migration: Migrated from Pinecone to MongoDB Atlas Vector Search for improved performance and cost efficiency - Word Statistics Backend: Enhanced server-side tracking of writing metrics for paid users - Settings Page UX: Complete redesign with tabbed interface for better organization - History Sidebar: Improved readability ...

... things I should consider? Comment from Sanskar Yadav(@sanskarix): Although I switch between tools and walkarounds a lot but here is my current stack: Python (for quick prototyping) FastAPI Node (for integrations and automations) Vector DBs like Pinecone or Postgres+pgvector Notion for docs and content workflows Sometimes Rust for speed - critical stuff I stopped using a bunch of all-in-one platforms - too heavy and not flexible enough. Sticking to modular tools that do one thing well ... ... easy to swap out! Comment from Lakeesha Weatherwax(@deleted-9027130): Right now, I m vibing with Rust for performance-heavy parts, Python for prototyping, and Pinecone

... just get the free license. But, if you build something substantial then you're into a paid account. TL;DR too many options and licenses. I use the following tools consistently: OpenAI (API and Chat) Claude Gemini n8n Pinecone Google Search API Google Workplace Descript Random signups that I probably won't keep Airtable Perplexity Zapier Lovable (I do like this one) Bedrock Comment from Matthew @ Sapling(@tinyorgtech): @busmark_w_nika  no. In fact I'm going through them ... ... Howell(@howell4change): A few but not a complete list and @OpenAI @Figma @Gamma (super useful) @Fitbod Weight Lifting Workout (a truly great product) @usebubbles , can't live without, I like the permissions and the way the notes are presented vs

... shipping a bunch of products recently, but after ~1 week of building, I recently launched (first launch) an MVP for AIBillingDashboard.com . What I built: A unified dashboard that tracks spending across all major AI providers (OpenAI, Claude, Gemini, Tavily, Pinecone, Make.com , n8n, HuggingFace, etc). Cost comparison tools that show which models give best value for specific use cases Usage analytics that identify cost-saving opportunities across your AI stack Billing cycle tracking and payment alerts to avoid surprise charges ... ... vector DBs, image gen. But at that point, why use a tool that only tracks AI usage separately? Wouldn't you want a tool that tracks all of your business spending, AI related or not? Who are your main competitors

... spending, getting surprised by bills), I decided to build. After ~1 week, I launched (first launch) an MVP for AIBillingDashboard.com . What I built: A unified dashboard that tracks spending across all major AI providers (OpenAI, Claude, Gemini, Tavily, Pinecone, Make.com , n8n, HuggingFace, etc). Cost comparison tools that show which models give best value for specific use cases Usage analytics that identify cost-saving opportunities across your AI stack Billing cycle tracking and payment alerts to avoid surprise charges ...

... Makers & Builders using LLMs! Anyone else building cool stuff like PDF Q&A or custom bots with RAG, but finding the context retrieval step... frustrating? Most AI app data stacks these days use vector search (Pinecone, Weaviate, etc.) to grab text chunks for the LLM. But sometimes it feels like it finds stuff that's keyword-similar while totally missing the actual point the user asked for. Leads to those slightly weak or "confidently wrong" LLM answers ...