The Group Chat
Helping people get their plans out of chat.
What works for everyone?
Let's make a plan.
What am I building?
A human first planning tool.
The Group Chat is a human first planning tool that lets people do more all while removing cognitive friction. Ultimately, I believe this piece of technology should help people feel more connected by helping them get their plans out of chat.
Why this problem?
It's a coordination problem.
Group travel planning isn't fundamentally an itinerary problem. It's a coordination problem between people with different preferences, constraints, communication styles, and levels of engagement. And honestly, I'm a planner and it solves a very real challenge I face often.
How I'm approaching it?
Support the conversation. Don't take it over.
I'm building The Group Chat around a simple principle: technology should support human connection without taking over the conversation. Instead of behaving like another person in the chat—or responding to every message—it observes what the group has shared, recognizes emerging preferences, questions, and tradeoffs, and contributes only when it can reduce coordination effort.
Sometimes the most useful response is intentional silence.
The product is being developed through small, testable slices. I document assumptions before implementing them, use the simplest architecture capable of testing the next hypothesis, and preserve ambiguity when the evidence does not yet justify a decision. Early experiments use transparent, deterministic behavior rather than prematurely introducing AI, recommendations, booking tools, accounts, analytics, or production-scale infrastructure.
The next learning step is a bounded test with a small group using their own devices. I'm testing whether joining feels natural, whether shared context remains understandable, whether facilitation helps without becoming intrusive, and whether the experience makes planning feel cognitively lighter. What I learn will determine what earns greater sophistication—and what should remain deliberately simple.
Where it is now
The foundational collaborative layer is live.
The foundational collaborative layer is live and undergoing its first external testing.
People can create an experience, invite others, join from their own devices, and contribute to a shared conversation. I'm now watching how real groups use that foundation: where coordination becomes difficult, whether shared context remains understandable, and whether the experience makes planning feel cognitively lighter.
I'm deliberately testing whether the human system works before asking AI to improve it.
What I learn here will determine what earns greater sophistication and what should remain deliberately simple.