Most reading does not stick. You finish an article, feel like you understood it, and then cannot explain it. SpeakOS fixes that loop: read short content, speak from memory, get AI feedback on clarity, structure, depth, and original thought.

It is not a fluency tool. It asks whether you actually understood what you read.

Flow: Select topic -> Read 5-minute article -> Speak without notes (1-3 min) -> Receive feedback + transcript.

Stack: React + Tailwind, Supabase, OpenAI Whisper, GPT analysis, and embeddings for article-vs-speech depth comparison.

Currently validating retention, learning impact, and daily usability.

A discovery engine for startup funding - grants, fellowships, and investor programs - aggregated and organized in one place.

The moat is the data infrastructure, not the UI.

300+ primary sources are monitored continuously. RSS feeds are scraped and parsed, then passed through an enrichment pipeline that tags opportunities by type, stage, geography, and eligibility. For sources without feeds, AI scrapers built on Notte and Browserless extract structured details including deadlines, amounts, criteria, and program specifics.

The result is a live database that stays current without manual curation.

100+ active grants tracked. 1000+ funding programs indexed.

Newsletter: 200+ subscribers, about 60% open rate, about 20% click rate, about 60% international readership.

Startup911 is the discovery layer. GrantSnap is the execution layer. Built to work together.

A Chrome extension that autofills grant and accelerator applications using structured founder data and AI.

Built solo end to end across two repos: a Next.js dashboard and a Chrome MV3 extension. AI layer on Gemini 3.1 Flash (primary) with Gemini 2.5 Flash fallback. Backend on Supabase with Deno Edge Functions. Billing live via Dodo Payments.

Ships with AI autofill, Deep Scan eligibility scoring, Magic Wands, Pitch Deck Builder, Pitch Script Generator, custom templates, advanced analytics, and data export.

Tiers: Explorer (free), Applicant ($59/mo), Operator ($99/mo).

Built at Acharya Ventures.

EERA is a central Founder OS: system of record + intelligence layer + interface layer.

Core context:

• Multi-workspace architecture with strict tenant isolation and role-based access
• Structured data backbone across teams, workflows, campaigns, leads, and knowledge
• AI layer for research, writing, analysis, and decision support grounded in internal context
• Knowledge base with RAG pipelines over workspace data and docs
• Modular apps for marketing, CRM/leads, operations, and execution visibility
• Integration layer for Meta, WhatsApp, logistics, email, and workflow triggers

Built solo in under 4 months to reduce founder tool fragmentation and decision fatigue.