{
 "name": "Vicaya",
 "tagline": "Build your knowledge base, prove it works, then ship it",
 "description": "Build your knowledge base, prove it works, then ship it.\n\nVicaya is a Mac app for developers adding retrieval-augmented generation (RAG) to an iOS app built with the Foundation Models framework. Drop in your documents, tune how they're split and searched, test the results against real questions, and export a ready-to-use knowledge bundle for Xcode. You don't need to be an ML specialist.\n\nSEE WHAT YOUR MODEL SEES\nAdd folders of Markdown, PDF and text files and preview every document as chunks with token counts. In the Playground, ask a question and watch the whole pipeline: which chunks were retrieved and their scores, the final prompt and its token count, the model's live answer, and how long each stage took.\n\nCHOOSE HOW KNOWLEDGE IS BUILT\n\nChunking: fixed size with overlap, by paragraph, or by heading\nIndex: Spotlight, keyword (full-text), vector (on-device embeddings), or hybrid\nContextual notes: the model adds a short note to each chunk describing where it sits in its document, so look-alike sections stay distinct\nTune chunk size, overlap, top-k and context budget to fit the on-device context window\n\nCHOOSE HOW QUESTIONS ARE ANSWERED\n\nDirect search with the question as asked\nQuery rewrite, which turns chatty follow-ups into good search queries\nTool mode, where the model decides when and what to search\nContext packing that removes duplicates, orders results, and fits them to your token budget\nVicaya only offers combinations that work together, and the Document Q&A template starts you with a sensible default.\n\nPROVE IT WORKS\nTesting is built in. Vicaya drafts test questions from your content, and you keep the good ones. Measure retrieval with recall and ranking metrics, then open any failing question to see what came back instead and why. Change one setting, rebuild in seconds, and see whether your scores went up.\n\nSHIP IT\nExport a single knowledge bundle and add it to your Xcode project. The included build plugin generates a typed Swift API with a ready-made search tool and a one-line answer call, powered by a lightweight runtime package. Your app runs exactly the pipeline you tested. The bundle records what it was built and tested with, so you're warned in debug builds when the device differs.\n\nBUILT FOR YOUR WORKFLOW\n\nOne project file holds your sources, settings and test set, and lives in your repo\nA command-line tool runs the same builds and evaluations in CI\nEvaluations use Apple's Evaluations framework, so they also run in Xcode\n\nGREAT FOR\n\nIn-app help centers and product manuals\nCourse, lesson and reference content\nAny app that ships known content and wants grounded, on-device answers\n\nRequires macOS 27. Exported bundles use iOS 27 features.",
 "url": "https://debuts.app/apps/vicaya",
 "category": "Utilities",
 "platforms": [
  "iPhone and iPad"
 ],
 "pricing": "free",
 "use_cases": [],
 "audience": [],
 "links": {
  "app_store": "https://apps.apple.com/us/app/vicaya/id6816322119?mt=12",
  "google_play": null,
  "website": "https://avitsadok.github.io/Vicaya-legal/"
 },
 "icon": "https://is1-ssl.mzstatic.com/image/thumb/Purple221/v4/78/7d/8e/787d8e86-a0da-c7b0-9a29-dddf12b4f06d/AppIcon-0-0-85-220-0-0-5-0-2x.png/512x512bb.png",
 "screenshots": [
  "https://is1-ssl.mzstatic.com/image/thumb/PurpleSource211/v4/a7/65/3d/a7653d65-6cdc-e03f-04b3-2d720f2b91ad/Screenshot_2026-09-26_at_8.58.26.png/800x500bb.jpg",
  "https://is1-ssl.mzstatic.com/image/thumb/PurpleSource221/v4/18/fa/50/18fa5074-f161-5d96-400d-c4acd6e42d57/Screenshot_2026-09-26_at_9.00.58.png/800x500bb.jpg",
  "https://is1-ssl.mzstatic.com/image/thumb/PurpleSource221/v4/c1/d4/ea/c1d4ea37-6d9e-baa2-26bb-efaf991fb22f/Screenshot_2026-09-26_at_9.01.59.png/800x500bb.jpg"
 ],
 "claimed": false,
 "verified": false,
 "maker": "Avi Tsadok",
 "upvotes": null,
 "followers": null,
 "rating": null,
 "ratings": 0,
 "awards": [],
 "launches": [],
 "ai_ready": 78,
 "badges": {
  "launched": "https://debuts.app/badge/vicaya?type=launched",
  "award": "https://debuts.app/badge/vicaya?type=award",
  "rating": "https://debuts.app/badge/vicaya?type=rating"
 }
}