OpenlySearch: Privacy-First Local AI Desktop Intelligence & Project Memory

Описание и применение

OpenlySearch – локальный десктопный помощник для профессионалов, обеспечивающий быстрый поиск, извлечение задач и планирование без передачи данных в облако.

Задача

Freelancers, developers, and consultants need AI assistance without violating NDAs or data residency laws, and they struggle with fragmented context across files, notes, and clipboard history.

Результат

OpenlySearch enables privacy‑first AI assistance for professionals, providing fast local search, task extraction, and calendar planning without external data transmission.

Рабочий процесс

• 1. Scan project folders and PDFs, generating 384‑dimensional embeddings with @huggingface/transformers.
• 2. Persist embeddings in a lightweight local index (userData/index.json).
• 3. Execute hybrid semantic + lexical search in under 5 ms.
• 4. Extract TODOs, FIXMEs, deadlines, and decisions via regex and optional local LLM (Ollama).
• 5. Structure deliverables into phases and bind tasks to an interactive monthly calendar with tick marks.
• 6. Monitor clipboard, filter out sensitive tokens, and store non‑sensitive snippets.
• 7. Generate daily memory briefings with text‑to‑speech audio readout.

Доказательства

OpenSearch turns passive local project folders into an active, actionable workspace

Источник

Every single feature runs on the user's hardware without sending a single byte to external servers

Источник

Hybrid Semantic + Lexical Search: Instant sub‑second search combining local MiniLM embeddings (all‑MiniLM‑L6‑v2) with keyword ranking over documents, PDFs, Markdown, and code

Источник

Searchable Clipboard Memory: Automatically captures and categorizes copied code, commands, and notes while proactively filtering out sensitive API keys and secrets

Источник