~/talk

What makes AI work in a hospital, and what makes it harder

Talkat WhatTheStack 2026 · 2026-09-14..09-19· [machine-learning]
links:none on file
slides:none on file

Real hospital AI needs deep integration: radiology voice recognition runs as two-stream real-time transcription tied to PACS and EHR so models can be fine-tuned with historical reports and recordings; GPT‑4o transcribe hits 100% on a full recording but fails as a live progressive stream, so post‑processing and personalization matter. OCR uses shadow-model swap (Maverick vs GPT‑4o) to handle drifting document types. Booking agents require patient index, code lists, scheduling and billing access; radiologists expect ~30 seconds per report.

AI-written summary · about

  1. 0:00Introduction
  2. 0:49Real-time voice recognition in radiology
  3. 9:55OCR for document scanning
  4. 13:23Call center AI agent integration
  5. 16:58AI for healthcare forms and workflows
  6. 18:41Conclusion and Q&A
~/search
type to search 6,518 conferences, 143,918 speakers and 192,133 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
← →
on a talk: 5 seconds back / forward
0–9
on a talk: jump to 0%–90% of the recording
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.