Hash is an 🏷️#ai-powered investment intelligence platform that transforms raw signal from across the internet into structured, evidence-weighted insight. It is built for investors, fund operators, and strategic decision-makers who need to see what's emerging before it's obvious — and act on it without drowning in noise.
What Hash Does
Hash operates a three-layer intelligence system:
- Landscape Tracking — Scrapes seven source types (Hacker News, Reddit, Substack, Product Hunt, GitHub, RSS, Indie Hackers) plus GDELT news monitoring on automated schedules. Raw content flows through a six-stage pipeline: ingestion → entity extraction → trend clustering → velocity measurement → builder identification → opportunity scoring
- Builder Monitoring — Within accelerating trends, Hash identifies the people and teams actively building. When a user adds an opportunity with GitHub, Substack, or RSS handles, Hash auto-provisions per-builder monitoring feeds that track their output — commits, releases, blog posts, star velocity — without requiring LLM calls
- Opportunity Assessment — Evidence-backed scoring with anti-inflation rules. Scores are capped at 40/100 with fewer than three evidence items and 60/100 if all evidence comes from a single source type. No speculation in summaries — only observed data
News Monitoring
Hash integrates GDELT-powered news event detection for real-time awareness of significant developments across the tech landscape. The system uses a dual-pass fetching strategy (article retrieval + body-text entity extraction), three-tier entity recognition, and an LLM validation gate that filters routine noise from genuinely significant events. News events surface on a dedicated dashboard with severity badges, expandable source articles, and direct links to create memos.
Content Creation Pipeline
Hash includes a hybrid content creation flow that bridges signal discovery with 📝MythOS knowledge capture. From any trend, builder, opportunity, or news event, a user can:
- Click "Create Memo" to open an AI-assisted drafting interface
- Optionally provide an angle or focus to guide the AI
- Review and edit the generated draft — title, markdown content, and tags
- Publish directly to their MythOS library with one click
Drafts are persisted independently of publishing, so edits are never lost even if MythOS is temporarily unavailable. The integration uses Sonnet for draft generation and requires a linked MythOS account (configurable in Preferences).
Daily Digest
Each user receives a daily email digest summarizing the top accelerating trends, highest-scored opportunities, and recent builder signals. The digest can optionally inject perspective from the user's MythOS library — surfacing what the user has previously written about the entities appearing in today's signal. Output format is configurable: text, audio, or combined.
Who Hash Is For
- Venture fund operators tracking emerging patterns before they become consensus. The primary user is Fred, who uses Hash to surface investment-grade opportunities with evidence trails
- Angel investors monitoring specific builders and technologies across multiple sources without manual RSS management
- Strategic operators who need structured awareness of what's moving in their industry — launches, funding rounds, regulatory shifts, crises — without managing a dozen tabs
Hash is not a deal sourcer. It detects landscape patterns first, then identifies individuals building within those patterns, then flags investment-worthy signals with evidence. The architecture enforces this discipline: trend-first, not company-first.
Hash is the synthesis engine — the layer that turns the internet's raw exhaust into structured thought. The content creation pipeline closes the loop between discovery and knowledge: what you find in Hash becomes memory in MythOS, and what you've written in MythOS shapes what Hash surfaces next. That feedback loop is the point.