Spaces:
Sleeping
Sleeping
Commit ·
5061a91
1
Parent(s): 30cc0d5
docs: drop scaffold-style "in two sentences" subheads in JUDGE_BRIEF
Browse files"## The problem, in two sentences" / "## What Sakhi does, in two sentences"
are prompt-instruction leakage as headers. A reader doesn't need to be
told the section is short. Renaming to plain topic headers.
- JUDGE_BRIEF.md +2 -2
JUDGE_BRIEF.md
CHANGED
|
@@ -2,11 +2,11 @@
|
|
| 2 |
|
| 3 |
*One-page version of the README. Full detail in [README.md](README.md). 3-min demo video: [youtu.be/n-u7J1lljUg](https://youtu.be/n-u7J1lljUg).*
|
| 4 |
|
| 5 |
-
##
|
| 6 |
|
| 7 |
India's 1 million+ ASHA health workers conduct 50M+ maternal and child home visits every year; every visit ends with a hand-filled paper form carried to the PHC. Danger signs observed in the field — preeclampsia, postpartum hemorrhage, neonatal distress — often don't reach the clinical system in time for intervention.
|
| 8 |
|
| 9 |
-
## What Sakhi does
|
| 10 |
|
| 11 |
Sakhi converts Hindi home-visit conversations (voice on a shared health-center workstation, text on the ASHA's phone offline) into structured NHM/MCTS forms + a function-calling-powered danger-sign triage that flags referrals with verbatim utterance evidence. Same pipeline, same anti-hallucination validation, two deployment modes: Whisper-Large + Gemma 4 E4B via Ollama on a workstation for accuracy, and Gemma 4 E2B via Cactus SDK on an Android phone for offline resilience.
|
| 12 |
|
|
|
|
| 2 |
|
| 3 |
*One-page version of the README. Full detail in [README.md](README.md). 3-min demo video: [youtu.be/n-u7J1lljUg](https://youtu.be/n-u7J1lljUg).*
|
| 4 |
|
| 5 |
+
## Problem
|
| 6 |
|
| 7 |
India's 1 million+ ASHA health workers conduct 50M+ maternal and child home visits every year; every visit ends with a hand-filled paper form carried to the PHC. Danger signs observed in the field — preeclampsia, postpartum hemorrhage, neonatal distress — often don't reach the clinical system in time for intervention.
|
| 8 |
|
| 9 |
+
## What Sakhi does
|
| 10 |
|
| 11 |
Sakhi converts Hindi home-visit conversations (voice on a shared health-center workstation, text on the ASHA's phone offline) into structured NHM/MCTS forms + a function-calling-powered danger-sign triage that flags referrals with verbatim utterance evidence. Same pipeline, same anti-hallucination validation, two deployment modes: Whisper-Large + Gemma 4 E4B via Ollama on a workstation for accuracy, and Gemma 4 E2B via Cactus SDK on an Android phone for offline resilience.
|
| 12 |
|