Pulse is a RAG-powered personal intelligence layer for founders. Paste your LinkedIn, GitHub, and a self-report once. Pulse builds a vector knowledge base of you. every chat retrieves your personal context before hitting the LLM. 4 background agents run silently: extracts commitments, tracks competitors daily via web search, parses reminders, drafts and schedules emails via Gmail. Multi-LLM: Gemini, Llama, Claude. switchable at runtime. Free to use. Open source.

Wispr Flow: Dictation That Works Everywhere — Stop typing. Start speaking. 4x faster.

Stop typing. Start speaking. 4x faster.

Hey PH 👋 This is my first RAG + agents project .. built it in 24 hours as a 3rd year CS student. A few things I'm genuinely proud of: - Embeddings run entirely in-browser via WASM (all-MiniLM-L6-v2). Zero API cost, works offline, cached after first load. No embedding vendor lock-in. - The competitor tracker runs a daily cron, fires 3 Serper searches per competitor, LLM-summarizes with urgency ratings, and injects high-urgency intel at the TOP of your next chat prompt. You don't check on competitors — they come to you. - 4 databases, each doing what it's best at: Qdrant for vectors, SQLite for open loops + intel, PostgreSQL for reminders + email actions, MongoDB for agent profiles. Would love feedback especially from founders who've tried tools like this before. What's missing? What would make you actually use it daily? Live → https://itspulse.vercel.app Code → https://github.com/herin7/pulse

Wispr Flow: Dictation That Works Everywhere — Stop typing. Start speaking. 4x faster.

Stop typing. Start speaking. 4x faster.

Pulse

Your AI co-founder that remembers everything

Your AI co-founder that remembers everything