Ruune by IoH
A MagSafe AI notetaker that solves the biggest failure in current AI: Capture Friction. Most AI tools only see what happens on your computer. Ruune captures the other 80% — coffee chats, phone calls, hallway meetings — with a single tap.
The Problem
Most AI tools only see what happens on your computer. But the most important conversations happen offline — in coffee chats, phone calls, and hallway meetings. If it's not a scheduled Zoom call, it doesn't exist.
The Product
Ruune is an ultra-thin (<3mm) MagSafe voice recorder that attaches to the back of your phone. By piggybacking on the one object you never leave behind, we reduce the friction of "remembering the device" to zero. This guarantees 100% data capture — extracting commitments, decisions, and open loops from your conversations and proactively surfacing them.
Core Insight
Standard notetakers treat memory as a search problem — retrieving static text from transcripts. We treat memory as a state problem — tracking the status of commitments, decisions, and relationships as they evolve over time. Instead of remembering to check, the system notifies you:
- → "You promised to email the deck by Friday."
- → "You haven't followed up with Mark about the roadmap you discussed last week."
- → "The DSO client is still waiting on the deliverables you committed to."
Competitive Landscape
Meeting Note Tools (Otter, Fireflies, Granola)
Event-based. Misses 80% of conversations that happen off-calendar. If it's not a scheduled Zoom call, it doesn't exist.
AI Voice Recorders (Plaud)
Static files. Great for capture, but no deep memory. They don't reason over your work or life.
Standalone AI Pendants (Limitless, Friend)
Form factor failure. Requiring a new daily habit is fatal. We piggyback on the phone you already carry.
Business Model
Self-hosted pipelines (Whisper + Pyannote on L4 GPUs) run at <$0.04/hr vs commodity APIs at ~$0.37/hr — a 10x cost advantage that secures 50%+ overall software margins.
Go-To-Market
Phase 1: Influencer-Driven (First 90 Days)
Direct relationships with professional creators (~$4K for 500K+ follower placements). Validating demand and collecting upfront cash before scaling spend.
Phase 2: Professional Adoption
Targeting knowledge workers (lawyers, doctors, founders, consultants) who lose money when context is lost. High willingness to pay and natural enterprise expansion paths.
Phase 3 — Enterprise Deployment
Organizations with 20+ employees in legal, consulting, and healthcare. Lead acquisition through existing relationships.
Moat & Defensibility
Data Network Effect
Memory value compounds over time
Switching Costs
High behavioral lock-in via temporal knowledge graph
Full-Stack Moat
Hardware + Software integration creates a deeper defensive layer than SaaS alone
Regulatory Barriers
Enterprise compliance (HIPAA/SOC2) experience creates entry barriers
The Ask
18 months runway to Series A. Capital allocated to Batch 1 Inventory, FCC Certification, Key Hires, and Brand Marketing.
Milestones:
- Ship Batch 1 (5k units)
- Achieve $1M ARR run rate
- Validate B2B expansion pilots
- Launch enterprise compliance tier
Team
Julian Zhang — CEO & Co-Founder
Former Google and Citadel engineer. Track record for rapid execution from prototype to product.
Joe Zeng — CTO & Co-Founder
Former Google and Citadel engineer. Founding engineer at YC-backed Atomicvest. Manufacturing relationships in Shenzhen.
Capture-only devices like Plaud are approaching ~$250M ARR with no intelligence layer. People are desperate to record their lives — we are building the intelligence layer that turns that data into value.