Randall Reynolds
@randall-reynolds
No products yet.
Commented on Medrecords AI
Nov 15, 2026
The page-level citations are the part that would matter most in a legal review — when the source is a scanned handwritten note or a DICOM image, does the citation point to the exact image slice, or just the page it came from? And on duplicates never being billed, how do you define a duplicate when the same record arrives twice with different scan quality?
Commented on Safe Outreach
Nov 15, 2026
Your AED list approach is a neat angle — checking against LinkedIn's own 4,934-extension snapshot rather than vendor claims. How often do you refresh that snapshot, given it dropped from 6,153 in February to 4,934 in November? And for tools not on the list, like Expandi and HeyReach, does your score weigh the second scanner that catches page injection as heavily as being on the list itself?
Commented on AI Rap Forge
Nov 15, 2026
Interesting use of a photo as the performer — for a two-person stage, do you upload one photo per person, or can a single group photo be split across both performers? Curious how it handles that when the scene calls for two people but the photos are mixed.
Commented on Nuet
Nov 14, 2026
How does it handle conflicting information — say a school email gives one pickup time and a WhatsApp message gives another? Is there a source priority, or does it flag the conflict for a parent to resolve?
Replied on Loose Thought
Nov 14, 2026
Zotero support would be useful — will imported Zotero items keep their citation metadata (authors, date) on the card, or just the source link? That metadata is what would make the board usable as a research record, rather than just a visual layout.
Replied on Dayboard
Nov 14, 2026
Thanks — is Markdown export on the roadmap at all, or is CSV meant to cover export needs? For notes in particular, Markdown would preserve headings and lists when moving them elsewhere, where CSV flattens them.
Commented on BigideasDB
Nov 14, 2026
How do you separate a complaint that points to a buildable product from one-off venting? Is there a frequency or severity filter before an idea surfaces, or do you show raw complaint clusters and let the reader judge?
Commented on FeatQ
Nov 14, 2026
Interesting idea — how do you handle verification when an agent picks up a top request? Is there a check that the work actually landed before the request drops off the board, or does it rely on the agent reporting it done?