PrivyPDF
PDF Q&A, export to Markdown, Word, JSON, and OCR workflows
Privacy-first AI tooling
Privy keeps high-context AI work close to the machine.
A local-first product family for documents, feeds, editing, API workflows, and reusable AI infrastructure.
local-first privacy-first kit-based multi-platform
Goals
Default to local custody.
- 1Keep sensitive documents off the cloud by default
- 2Reuse OCR, PDF, embedding, RAG, TTS, and related capabilities through a shared Kit layer
- 3Run one core stack across macOS, iOS, browser extensions, and optional desktop shells
Architecture
Shared kits, multiple product surfaces.
The product line is designed around reusable capability layers, so parsing, inference, providers, queues, and multimodal I/O do not need to be rebuilt per app.
Matrix
Products and kits
Each surface is a focused app, while the shared kit layer carries the heavy technical capability.
PrivyFeed
RSS aggregation, local/LAN inference summaries, and semantic search
PrivyIME
IME plus LLM augmentation with privacy and compliance boundaries
PrivyMermaid
Local Mermaid editing and preview across extension and native surfaces
PrivyApiStudio
API debugging, comparison, diff, and code generation tools
Kit layer
LlamaKit, DoclingKit, OCRKit, ImageKit, AudioKit, JobQueueKit, and ProviderKit