Inference Time Tactics is a new podcast from the Neurometric team focused on the emerging field of inference time compute. We explore how models reason, how infrastructure shapes outcomes, and what it takes to make systems reliable at runtime.
This series is for developers, researchers, and operators building the next generation of AI systems. Each episode looks at the real strategies behind runtime decision-making, inference orchestration, and performance tuning beyond the API call.
So far, we’ve covered:
Why inference time compute is becoming its own strategic surface
How developers are thinking about model configuration, cost, and reliability
What we’ve learned by replicating recent reasoning research
New episodes drop frequently, with companion posts and research updates shared here on Substack.
Watch and subscribe on all platforms:
Podbean: https://InferenceTimeTactics.podbean.com/e/episode-1/
Youtube: https://www.youtube.com/channel/UCKANQHWKbwf2_J3p6xgY3hA
Spotify: https://open.spotify.com/episode/2HvZPNvrLRShjQRDs8VzvJ?si=g4sRgIzvT5axQUaJEv2RNA
Rob May Calvin Cooper Byron Galbraith Dave Rauchwerk