Paper works because it's single-channel. No notifications, no tabs, no competing inputs. BrainDump is built on that same principle. You see one card. Not a list. Not a dashboard. One task, full screen. The only difference is the input is your voice instead of a pen, which matters when the executive function barrier is the act of writing things down in the first place.
Impressive demo. Curious how your TTS + STT + LLM stack handles real-time function calling and sustained phone/WebRTC sessions. Any benchmarks on latency or call quality?
Not everyone is charging by the minute, we're looking at a flat fee based on the call intent / type, for example in Property management: Call to Tour / Schedule Showing – prospective tenant wants to view a unit
New Lease / Application – questions or starting an application
Maintenance Request – routine repairs, appliance issues, etc.