Spaces:
Running
Running
File size: 11,626 Bytes
e954acb b384e24 e954acb |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 |
from modules.image_analysis import get_llm
from configs.load_config import PROMPT_LIBRARY, APP_STEPS_CONFIGS
from pydantic import BaseModel
from typing import Optional, Literal, List, Dict, Any
import json
import random
import time
class DateLocation(BaseModel):
date: Optional[str] = None
time: Optional[str] = None
location: Optional[str] = None
class PartiesInvolved(BaseModel):
other_driver_name: Optional[str] = None
other_driver_vehicle: Optional[str] = None
witnesses: Optional[str] = None
class FaultAssessment(BaseModel):
who_at_fault: Literal["me", "other_driver", "unclear"]
reason: str
class IncidentDescription(BaseModel):
what_happened: str
damage_severity: Literal["minor", "moderate", "severe", "unclear"]
class InjuriesMedical(BaseModel):
anyone_injured: Literal["yes", "no", "unknown"]
injury_details: Optional[str] = None
medical_attention: Optional[str] = None
injury_severity: Optional[Literal["none", "minor", "moderate", "severe", "unclear"]]
class FunctionCallResult(BaseModel):
function_name: str
result: Dict[str, Any]
status: Literal["success", "error"]
message: str
class IncidentAnalysis(BaseModel):
date_location: DateLocation
parties_involved: PartiesInvolved
fault_assessment: FaultAssessment
incident_description: IncidentDescription
injuries_medical: InjuriesMedical
function_calls_made: List[FunctionCallResult] = []
external_data_retrieved: Dict[str, Any] = {}
def mock_weather_lookup(date: str, location: str) -> Dict[str, Any]:
"""Mock function to look up weather conditions for a specific date and location"""
time.sleep(0.5) # Simulate API call delay
weather_conditions = [
"Clear",
"Rainy",
"Foggy",
"Snowy",
"Overcast",
"Partly Cloudy",
]
temperatures = range(20, 85)
return {
"date": date,
"location": location,
"temperature": f"{random.choice(temperatures)}°F",
"conditions": random.choice(weather_conditions),
"visibility": random.choice(["Good", "Poor", "Fair"]),
"precipitation": random.choice(["None", "Light Rain", "Heavy Rain", "Snow"]),
"wind_speed": f"{random.randint(0, 25)} mph",
}
def mock_driver_record_check(
driver_name: str, license_plate: str = None
) -> Dict[str, Any]:
"""Mock function to check driver record and vehicle registration"""
risk_levels = ["Low", "Medium", "High"]
return {
"driver_found": True,
"license_status": random.choice(["Valid", "Suspended", "Expired"]),
"insurance_status": random.choice(["Active", "Lapsed", "Unknown"]),
"previous_claims": random.randint(0, 5),
"violations_last_3_years": random.randint(0, 3),
"risk_assessment": random.choice(risk_levels),
"vehicle_registration": "Valid" if license_plate else "Not checked",
}
AVAILABLE_FUNCTIONS = {
"weather_lookup": {
"name": "weather_lookup",
"description": "Look up weather conditions for a specific date and location to understand incident context",
"parameters": {
"type": "object",
"properties": {
"date": {"type": "string", "description": "Date of the incident"},
"location": {
"type": "string",
"description": "Location where incident occurred",
},
},
"required": ["date", "location"],
},
"function": mock_weather_lookup,
},
"driver_record_check": {
"name": "driver_record_check",
"description": "Check driving record and insurance status of other party involved",
"parameters": {
"type": "object",
"properties": {
"driver_name": {
"type": "string",
"description": "Name of the other driver",
},
"license_plate": {
"type": "string",
"description": "License plate number if available",
},
},
"required": ["driver_name"],
},
"function": mock_driver_record_check,
},
}
def execute_function_call(
function_name: str, parameters: Dict[str, Any]
) -> FunctionCallResult:
"""Execute a function call and return the result"""
try:
if function_name not in AVAILABLE_FUNCTIONS:
return FunctionCallResult(
function_name=function_name,
result={},
status="error",
message=f"Function {function_name} not found",
)
function_impl = AVAILABLE_FUNCTIONS[function_name]["function"]
result = function_impl(**parameters)
return FunctionCallResult(
function_name=function_name,
result=result,
status="success",
message=f"Successfully executed {function_name}",
)
except Exception as e:
return FunctionCallResult(
function_name=function_name,
result={},
status="error",
message=f"Error executing {function_name}: {str(e)}",
)
def process_transcript_description(transcript: str, api_key: str):
"""
Analyze the provided transcript and extract structured information for insurance claim processing.
Now includes function calling capabilities to gather additional context.
Args:
transcript: transcript string to process
api_key: api key to use
Returns:
incident_description: incident description with function call results
"""
print("Starting incident analysis with function calling...")
llm = get_llm(
api_key=api_key,
model="accounts/fireworks/models/llama4-scout-instruct-basic",
temperature=APP_STEPS_CONFIGS.incident_response.temperature,
)
# Enhanced prompt that includes function calling
prompt_text = f"""
{PROMPT_LIBRARY["incident_processing"]["advanced"]}
**ADDITIONAL CAPABILITIES:**
You now have access to external functions that can help gather additional context for the claim.
Based on the transcript, you should consider calling these functions if the information would be helpful:
Available Functions:
{json.dumps([{k: {kk: vv for kk, vv in v.items() if kk != 'function'}} for k, v in AVAILABLE_FUNCTIONS.items()], indent=2)}
**IMPORTANT:**
1. First, extract all the basic incident information from the transcript
2. Then determine which functions (if any) would provide helpful additional context
3. For each function you want to call, include a "function_calls" section in your response
4. I will execute the functions and provide you with the results
5. You will then incorporate those results into your final analysis
**TRANSCRIPT TO ANALYZE:**
<transcript>
{transcript}
</transcript>
Please provide your initial analysis and specify any function calls you'd like to make.
"""
# First pass - get initial analysis and any function calls
response = llm.chat.completions.create(
messages=[
{
"role": "system",
"content": "You are an expert automotive claims adjuster analyzing vehicle damage with access to external data sources.",
},
{"role": "user", "content": prompt_text},
],
response_format={
"type": "json_object",
"schema": IncidentAnalysis.model_json_schema(),
},
temperature=APP_STEPS_CONFIGS.incident_response.temperature,
)
# Parse initial response
incident_data = IncidentAnalysis.model_validate_json(
response.choices[0].message.content
)
print("Initial analysis complete. Checking for function calls...")
function_calls_to_make = []
external_data = {}
# Making it more robust by checking for necessary inputs
if (
incident_data.date_location.date
and incident_data.date_location.location
and incident_data.date_location.date.lower()
not in ["not specified", "unknown", ""]
):
function_calls_to_make.append(
{
"name": "weather_lookup",
"params": {
"date": incident_data.date_location.date,
"location": incident_data.date_location.location,
},
}
)
if (
incident_data.parties_involved.other_driver_name
and incident_data.parties_involved.other_driver_name.lower()
not in ["information not provided", "not specified", ""]
):
function_calls_to_make.append(
{
"name": "driver_record_check",
"params": {
"driver_name": incident_data.parties_involved.other_driver_name,
"license_plate": incident_data.parties_involved.other_driver_vehicle,
},
}
)
# Execute tool calls
function_results = []
if function_calls_to_make:
print(f"Executing {len(function_calls_to_make)} function calls...")
for call in function_calls_to_make:
print(f" - Calling {call['name']}...")
result = execute_function_call(call["name"], call["params"])
function_results.append(result)
if result.status == "success":
external_data[call["name"]] = result.result
incident_data.function_calls_made = function_results
incident_data.external_data_retrieved = external_data
# Update analysis with external data pulled from tools
if function_results:
print("Incorporating external data into final analysis...")
enhancement_prompt = f"""
Based on the initial incident analysis and the additional data retrieved from external sources,
please provide an enhanced analysis that incorporates this new information.
**INITIAL ANALYSIS:**
{incident_data.model_dump_json(indent=2)}
**EXTERNAL DATA RETRIEVED:**
{json.dumps(external_data, indent=2)}
Please update your analysis to incorporate insights from this external data where relevant.
For example:
- Weather conditions might affect fault assessment
- Other traffic incidents might provide context
- Driver record might influence risk assessment
- Medical facility info might be relevant for injury cases
Provide the complete updated analysis.
"""
enhanced_response = llm.chat.completions.create(
messages=[
{
"role": "system",
"content": "You are an expert automotive claims adjuster incorporating external data into your analysis.",
},
{"role": "user", "content": enhancement_prompt},
],
response_format={
"type": "json_object",
"schema": IncidentAnalysis.model_json_schema(),
},
temperature=APP_STEPS_CONFIGS.incident_response.temperature,
)
enhanced_data = IncidentAnalysis.model_validate_json(
enhanced_response.choices[0].message.content
)
enhanced_data.function_calls_made = function_results
enhanced_data.external_data_retrieved = external_data
incident_data = enhanced_data
print("Finished incident analysis with function calling.")
return incident_data
|