[Ryan — the one line a panel should take from twenty years on the tools]
Five-camera crew (staff + hired freelancers), coordinated on-site. Full night cut to sixty seconds.
MuchMusic. [Ryan — one-line context if worth naming]
CTV. [Ryan — one-line context]
CTV. [Ryan — one-line context]
MuchMusic. [Ryan — one-line context]
CTV. [Ryan — one-line context]
MTV / MuchMusic. [Ryan — one-line context]
[Ryan — what running Central Video actually means: team size, cadence, output]
Canadian Business. [Ryan — what the piece is, live/pre/post]
[Brand]. [Ryan — piece description]
[Brand]. [Ryan — piece description]
[Ryan — the thesis: where AI earns its place in a real production pipeline]
[Ryan — why phone-camera AR on static signage matters commercially]
Point a phone at the TTC ad. Tracking frames the ad. The CP24 helicopter flies out into the platform and returns to the ad.
Same pattern works on any static surface — bus shelters, in-store signage, print, endcaps.
TSN. [Ryan — what it was, where it lived, what the interaction was]
CP24. [Ryan — description]
MuchMusic. [Ryan — description]
CTV. [Ryan — description]
[Ryan — what these two tools changed for the team day to day]
[Ryan — 1-2 sentences: what it does, what pain it solves]
[Ryan — 1-2 sentences: what it does, where deployed]
[Ryan — what being CTO of a clinical AI product demonstrates for this role]
Innermap builds Affi, an EFT-informed AI companion. I lead product, engineering, and evals. Katherine Prokopakis (registered psychotherapist) leads clinical.
Everything user-facing passes through the evaluation pipeline before it ships. See CH.07.
[Ryan — why testing is the discipline that actually transfers]
AI shifts one discipline more than any other: testing. What did the model produce, did it drift, does the rubric still catch what it should. Taste and craft don't change; QA does.
Same shape works for brand content as for a therapy companion. Versioned prompts. Rubric per client. Automated regression. Human sampling on the edges.