Systems thinking for human interactions
March 8th, 2025
Jason George, Piotr Kram & Team
Try our demo
Think of consumer interactions in two modes:
nn
Deterministic interactions which follow pre-defined journeys (e.g. supermarket check-out)
Non-deterministic interactions which allow for relatively undefined scenarios
Previous wave software has been automating '1.'.
Current wave software is eating '2.'.
Previous wave software has been automating human interactions.
Current wave software is humanising digital interactions.
We build systems which deliver on the promise of digital-human interactions.
Good companies have been, automating low-consequence, deterministic interactions (bottom-left tail) for a long time.
The problem is, these automations are shallow and highly rigid. And the human experiences left behind are often no better.
As LLM agents allow us to now automate non-deterministic, 'human' interactions (top-left tail), the opportunity we see is for great companies to retain the richness of interaction across both extremes.
The same underlying technology that allows an LLM to speak to you about your goals/feelings, could be used to help humans do this more sensitively; or to help other LLMs with the same.
When the world moves to build vertical agents solving discrete 'problems'. We build integrated systems that transform the whole experience.
This is our moat.
