Demo Build · Knowledge Assistant · RAG
An AI assistant for your knowledge — on Telegram, by text or voice
A "second brain" I built for myself: ask questions across a Google Workspace Shared Drive and an Obsidian vault, straight from Telegram — typed or spoken. Every answer is grounded in my own notes and documents. The same system points at a team's sources.
RAGClaudeVector searchGoogle WorkspaceObsidianTelegramVoice (speech-to-text)Python
Architecture — Google Drive & Obsidian → RAG index → Telegram (text & voice), with the source behind every answer.
The challenge
Useful knowledge ends up scattered — documents and spreadsheets in Google Drive, Markdown notes in Obsidian, things you wrote down somewhere months ago. Finding the right answer means digging across tools on a laptop, when half the time the question pops into your head on the move.
What I built
An AI assistant that indexes both a Google Workspace Shared Drive and an Obsidian vault, and answers questions over all of it from a Telegram chat — typed or spoken. I run it as a personal "second brain"; the same pipeline points at a team's own sources and channel.
- Two sources, one brain: ingests a Google Shared Drive (Docs, PDFs, Sheets) and an Obsidian Markdown vault into a single retrieval layer (RAG).
- Ask by voice: send a Telegram voice note and it's transcribed (speech-to-text), then answered like any typed question.
- Grounded answers: replies are built from your own documents — with the note or file behind each one.
- Reachable anywhere: it lives in Telegram — no new app to install, and it works from your phone.
How it works
- Syncs and chunks documents from the Shared Drive and the Obsidian vault, embedding them into a vector index that stays current as files change.
- A Telegram bot receives each message; voice notes are transcribed to text first.
- Retrieves the most relevant passages across both sources and answers from that context with Claude.
- Replies in the chat with the answer and the note or document it drew on.
Why it matters
- Your team's knowledge — Drive, docs, notes, wikis — becomes answerable in plain language.
- Reachable from a tool people already keep open: Telegram (or Slack / WhatsApp), by text or voice.
- Grounded in your own content, so answers stay accurate and traceable.
- Points at your sources and goes live in about two weeks.
Tech
Retrieval-Augmented GenerationClaude APIEmbeddings / vector searchGoogle Workspace API (Shared Drive)Obsidian (Markdown)Telegram Bot APISpeech-to-text (voice)Python
Built for my own knowledge base (Google Drive + Obsidian). The same pipeline points at your sources — Drive, Notion, Confluence, a docs folder — and the channel your team already uses.
Want your team's knowledge one message away?
I can wire this to your sources and channels in about two weeks. Just get in touch.
Get in touch →