Systems that
understand people

Getting hands dirty with the tools I evaluate. Voice agents, avatar platforms, AI concierges. Dispatches from trying the thing, not just reading about it.

Part of how I think about frontier AI is by building with it. These are active experiments where I'm testing how well conversational AI actually works when it meets real people in real contexts. Not demos. Deployed systems, collecting signal.

Active experiments
Active
Aptiq
Can an AI avatar conduct a structured interview that surfaces real candidate signal? Testing whether conversational pre-screening beats form-based filtering for high-volume hiring.
avatar interviews voice AI HR screening
See the experiment
Active
Cadence
What happens when hotel guests can just talk to a voice agent instead of calling the front desk? Testing multilingual voice AI as a concierge layer in hospitality.
voice concierge multilingual hospitality
See the experiment
Building
Elen
Can a voice AI sparring partner actually improve how you think? Daily 20-minute sessions that pressure-test your reasoning across five workout types, with adaptive difficulty and skill tracking.
cognitive training voice AI critical thinking
See the experiment