File size: 7,369 Bytes
0af0679
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### llm_legal_advisor (Parallelization-pattern)\n",
    "\n",
    "#### Overview\n",
    "This module implements a parallel legal document analysis system using multiple AI agents to process legal documents concurrently."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Start with imports \n",
    "import os\n",
    "import json\n",
    "from dotenv import load_dotenv\n",
    "from openai import OpenAI\n",
    "from IPython.display import Markdown, display\n",
    "import concurrent.futures"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "load_dotenv(override=True)\n",
    "open_api_key = os.getenv(\"OPENAI_API_KEY\")\n",
    "groq_api_key = os.getenv(\"GROQ_API_KEY\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Helper Functions\n",
    "\n",
    "##### Technical Details\n",
    "- **Concurrency**: Uses ThreadPoolExecutor for parallel processing\n",
    "- **API**: Groq API with OpenAI-compatible interface\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### `llm_summarizer`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Summarizes legal documents using AI\n",
    "def llm_summarizer(document: str) -> str:\n",
    "    response = OpenAI(api_key=groq_api_key, base_url=\"https://api.groq.com/openai/v1\").chat.completions.create(\n",
    "        model=\"llama-3.3-70b-versatile\",\n",
    "        messages=[\n",
    "            {\"role\": \"system\", \"content\": \"You are a corporate lawyer. Summarize the key points of legal documents clearly.\"},\n",
    "            {\"role\": \"user\", \"content\": f\"Summarize this document:\\n\\n{document}\"}\n",
    "        ],\n",
    "        temperature=0.3,\n",
    "    )\n",
    "    return response.choices[0].message.content"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### `llm_evaluate_risks`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Identifies and analyzes legal risks in documents\n",
    "def llm_evaluate_risks(document: str) -> str:\n",
    "    response = OpenAI(api_key=groq_api_key, base_url=\"https://api.groq.com/openai/v1\").chat.completions.create(\n",
    "        model=\"llama-3.3-70b-versatile\",\n",
    "        messages=[\n",
    "            {\"role\": \"system\", \"content\": \"You are a corporate lawyer. Identify and explain legal risks in the following document.\"},\n",
    "            {\"role\": \"user\", \"content\": f\"Analyze the legal risks:\\n\\n{document}\"}\n",
    "        ],\n",
    "        temperature=0.3,\n",
    "    )\n",
    "    return response.choices[0].message.content"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### `llm_tag_clauses`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Classifies and tags legal clauses by category\n",
    "def llm_tag_clauses(document: str) -> str:\n",
    "    response = OpenAI(api_key=groq_api_key, base_url=\"https://api.groq.com/openai/v1\").chat.completions.create(\n",
    "        model=\"llama-3.3-70b-versatile\",\n",
    "        messages=[\n",
    "            {\"role\": \"system\", \"content\": \"You are a legal clause classifier. Tag each clause with relevant legal and compliance categories.\"},\n",
    "            {\"role\": \"user\", \"content\": f\"Classify and tag clauses in this document:\\n\\n{document}\"}\n",
    "        ],\n",
    "        temperature=0.3,\n",
    "    )\n",
    "    return response.choices[0].message.content"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### `aggregator`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Organizes and formats multiple AI responses into a structured report\n",
    "def aggregator(responses: list[str]) -> str:\n",
    "    sections = {\n",
    "        \"summary\": \"[Section 1: Summary]\",\n",
    "        \"risk\": \"[Section 2: Risk Analysis]\",\n",
    "        \"clauses\": \"[Section 3: Clause Classification & Compliance Tags]\"\n",
    "    }\n",
    "\n",
    "    ordered = {\n",
    "        \"summary\": None,\n",
    "        \"risk\": None,\n",
    "        \"clauses\": None\n",
    "    }\n",
    "\n",
    "    for r in responses:\n",
    "        content = r.lower()\n",
    "        if any(keyword in content for keyword in [\"summary\", \"[summary]\"]):\n",
    "            ordered[\"summary\"] = r\n",
    "        elif any(keyword in content for keyword in [\"risk\", \"liability\"]):\n",
    "            ordered[\"risk\"] = r\n",
    "        else:\n",
    "            ordered[\"clauses\"] = r\n",
    "\n",
    "    report_sections = [\n",
    "        f\"{sections[key]}\\n{value.strip()}\"\n",
    "        for key, value in ordered.items() if value\n",
    "    ]\n",
    "\n",
    "    return \"\\n\\n\".join(report_sections)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### `coordinator`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Orchestrates parallel execution of all legal analysis agents\n",
    "def coordinator(document: str) -> str:\n",
    "    \"\"\"Dispatch document to agents and aggregate results\"\"\"\n",
    "    agents = [llm_summarizer, llm_evaluate_risks, llm_tag_clauses]\n",
    "    with concurrent.futures.ThreadPoolExecutor() as executor:\n",
    "        futures = [executor.submit(agent, document) for agent in agents]\n",
    "        results = [f.result() for f in concurrent.futures.as_completed(futures)]\n",
    "    return aggregator(results)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<b>Lets ask our legal corporate advisor</b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "dummy_document = \"\"\"\n",
    "This agreement is made between ABC Corp and XYZ Ltd. The responsibilities of each party shall be determined as the project progresses.\n",
    "ABC Corp may terminate the contract at its discretion. No specific provisions are mentioned regarding data protection or compliance with GDPR.\n",
    "For more information, refer the clauses 10 of the agreement.\n",
    "\"\"\"\n",
    "\n",
    "final_report = coordinator(dummy_document)\n",
    "print(final_report)\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}