Hugging Face's logo

Spaces:
SRVCP
/
mrAgent
Running

App Files Community
Rakesh commited on
Commit
2c278e8
·
verified ·
1 Parent(s): b0eaf65

Update index.html

Browse files
Files changed (1) hide show
  1. index.html +156 -18
index.html CHANGED
@@ -1,19 +1,157 @@
1
- <!doctype html>
2
- <html>
3
- <head>
4
- <meta charset="utf-8" />
5
- <meta name="viewport" content="width=device-width" />
6
- <title>My static Space</title>
7
- <link rel="stylesheet" href="style.css" />
8
- </head>
9
- <body>
10
- <div class="card">
11
- <h1>Welcome to your static Space!</h1>
12
- <p>You can modify this app directly by editing <i>index.html</i> in the Files and versions tab.</p>
13
- <p>
14
- Also don't forget to check the
15
- <a href="https://huggingface.co/docs/hub/spaces" target="_blank">Spaces documentation</a>.
16
- </p>
17
- </div>
18
- </body>
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
19
  </html>
 
1
+ <!DOCTYPE html>
2
+ <html lang="en">
3
+ <head>
4
+ <meta charset="UTF-8" />
5
+ <meta name="viewport" content="width=device-width, initial-scale=1.0" />
6
+ <title>Medical Research Architecture</title>
7
+ <style>
8
+ body {
9
+ font-family: Arial, sans-serif;
10
+ background: #f7f9fa;
11
+ margin: 0;
12
+ padding: 40px;
13
+ }
14
+ .step {
15
+ background: #ffffff;
16
+ border-left: 6px solid #4CAF50;
17
+ padding: 20px;
18
+ margin-bottom: 30px;
19
+ border-radius: 10px;
20
+ box-shadow: 0 2px 6px rgba(0, 0, 0, 0.1);
21
+ }
22
+ .step h2 {
23
+ margin-top: 0;
24
+ }
25
+ .agents {
26
+ display: flex;
27
+ flex-wrap: wrap;
28
+ gap: 20px;
29
+ margin-top: 15px;
30
+ }
31
+ .agent-box {
32
+ background: #e9f5ff;
33
+ border: 1px solid #cce7ff;
34
+ border-radius: 8px;
35
+ padding: 10px 15px;
36
+ flex: 1 1 30%;
37
+ min-width: 250px;
38
+ box-shadow: 0 1px 4px rgba(0, 0, 0, 0.05);
39
+ }
40
+ .arrow {
41
+ text-align: center;
42
+ margin: 10px 0;
43
+ font-size: 20px;
44
+ }
45
+ .note {
46
+ font-style: italic;
47
+ font-size: 0.9em;
48
+ color: #555;
49
+ margin-top: 10px;
50
+ }
51
+ </style>
52
+ </head>
53
+ <body>
54
+ <div class="step">
55
+ <h2>STEP 1: Parallelized Data Collection</h2>
56
+ <p><strong>Supervisor Agent</strong> dispatches specialized agents to fetch relevant data in parallel.</p>
57
+ <div class="agents">
58
+ <div class="agent-box">
59
+ <strong>GBQ Research Agent</strong><br/>
60
+ - Fetches structured internal datasets from BigQuery.
61
+ </div>
62
+ <div class="agent-box">
63
+ <strong>DailyMed Agent</strong><br/>
64
+ - Retrieves verified drug information and monographs.
65
+ </div>
66
+ <div class="agent-box">
67
+ <strong>Web Research Agent</strong><br/>
68
+ - Crawls and scrapes open medical research and news.
69
+ </div>
70
+ </div>
71
+ </div>
72
+
73
+ <div class="step">
74
+ <h2>STEP 2: Data Synthesis & Gap Analysis</h2>
75
+ <p>The <strong>Synthesizer Agent</strong> merges the raw findings into a preliminary brief, resolving conflicts and identifying gaps.</p>
76
+ <div class="agents">
77
+ <div class="agent-box">
78
+ <strong>Input</strong><br/>
79
+ - Data from all Step 1 Agents.
80
+ </div>
81
+ <div class="arrow">⟶</div>
82
+ <div class="agent-box">
83
+ <strong>Synthesizer Agent</strong><br/>
84
+ - Creates: <em>Preliminary Research Brief</em>
85
+ <div class="note">Highlights inconsistencies and missing data points.</div>
86
+ </div>
87
+ </div>
88
+ </div>
89
+
90
+ <div class="step">
91
+ <h2>STEP 3: Collaborative Summarization & Voting</h2>
92
+ <p>Multiple LLM agents draft independent summaries, then vote to reach a unified consensus.</p>
93
+ <div class="agents">
94
+ <div class="agent-box">
95
+ <strong>Summarizer Agent A</strong><br/>
96
+ - Powered by Claude<br/>
97
+ - Drafts version 1
98
+ </div>
99
+ <div class="agent-box">
100
+ <strong>Summarizer Agent B</strong><br/>
101
+ - Powered by Gemini<br/>
102
+ - Drafts version 2
103
+ </div>
104
+ <div class="agent-box">
105
+ <strong>Summarizer Agent C</strong><br/>
106
+ - Powered by GPT-4o<br/>
107
+ - Drafts version 3
108
+ </div>
109
+ </div>
110
+ <div class="arrow">⟶</div>
111
+ <div class="agents">
112
+ <div class="agent-box">
113
+ <strong>Critique & Voting Agent</strong><br/>
114
+ - Compares drafts against a rubric<br/>
115
+ - Creates: <em>Consensus Summary</em>
116
+ <div class="note">Integrates strengths from all drafts.</div>
117
+ </div>
118
+ </div>
119
+ </div>
120
+
121
+ <div class="step">
122
+ <h2>STEP 4: Human-in-the-Loop Review</h2>
123
+ <p>Medical researchers review AI-generated content, provide feedback, and request clarification.</p>
124
+ <div class="agents">
125
+ <div class="agent-box">
126
+ <strong>Interface Viewer</strong><br/>
127
+ - Displays Consensus Summary and source info<br/>
128
+ - Enables targeted review by human experts
129
+ </div>
130
+ <div class="arrow">⟶</div>
131
+ <div class="agent-box">
132
+ <strong>Medical Researcher (MR)</strong><br/>
133
+ - Reviews key chunks<br/>
134
+ - Provides corrections, queries, and follow-up requests
135
+ </div>
136
+ </div>
137
+ </div>
138
+
139
+ <div class="step">
140
+ <h2>STEP 5: Final Report Generation</h2>
141
+ <p>The final LLM synthesizes validated data into a formal, structured report.</p>
142
+ <div class="agents">
143
+ <div class="agent-box">
144
+ <strong>Report Generation Agent</strong><br/>
145
+ - Powered by GPT-4o / Gemini 1.5 Pro<br/>
146
+ - Inputs: raw data, consensus summary, MR feedback
147
+ </div>
148
+ <div class="arrow">⟶</div>
149
+ <div class="agent-box">
150
+ <strong>Final Research Document</strong><br/>
151
+ - Fully structured, validated, and ready for submission or publication
152
+ </div>
153
+ </div>
154
+ </div>
155
+
156
+ </body>
157
  </html>