Who I am:
I’m Teddy Williams — founder/operator building Obsidian Star with my business partner CJ. I work across hospitality, business development, AI systems, real estate, investing, branding, and long-term personal knowledge management.
How I’ll use Pocket:
I’m using Pocket as a memory layer / Chief of Staff for real conversations: business calls, strategy sessions, social conversations, family discussions, idea capture, relationship tracking, and long-term pattern recognition. My goal is not just transcription — I want Pocket to help me remember what mattered, who said what, what changed, what needs follow-up, and what may happen next.
My setup:
I’m currently using Pocket with ChatGPT/Claude, Google Drive, and Obsidian as part of a larger “external brain” system. I’m testing Auto Detect vs. custom templates, especially for summaries that need to preserve speaker attribution, action items, relationship intelligence, and predictions without becoming too long.
One question:
What’s the best way to understand how Pocket interprets custom summary templates? Specifically, does it treat each section as mandatory, or can prompts make sections more adaptive depending on the conversation type?
One request:
I’d love a mode or setting for speaker/entity review before summarization — something like a quick “cast list” where I can confirm: active speakers, background voices, and people merely discussed. That would make relationship tracking and long-term memory much more accurate.