Spaces:
Build error
Build error
File size: 34,786 Bytes
77248cd | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 | """
Headroom Worst-Case Benchmark: Where Compression Hurts
This benchmark tests scenarios where Headroom's statistical compression
may NOT be beneficial - to understand the limits of the approach.
Worst cases for Headroom:
1. Highly unique data (no patterns to compress)
2. Data where every item is equally important
3. Data where subtle differences matter
4. Small datasets (not enough data for statistics)
5. Data where you need EXACT recall (audit/legal)
"""
import hashlib
import json
import os
import random
import time
from dataclasses import dataclass
from typing import Any
try:
from openai import OpenAI # noqa: F401
OPENAI_AVAILABLE = True
except ImportError:
OPENAI_AVAILABLE = False
try:
from headroom import HeadroomClient, OpenAIProvider
HEADROOM_AVAILABLE = True
except ImportError:
HEADROOM_AVAILABLE = False
# =============================================================================
# WORST-CASE DATA GENERATORS
# =============================================================================
def generate_unique_support_tickets(num_tickets: int = 50) -> dict:
"""
Customer support tickets where EVERY ticket is unique and important.
No redundancy - each customer has a different problem.
This is hard for Headroom because:
- No repeated patterns to compress
- Every ticket needs attention
- Can't safely remove any ticket
"""
products = ["Pro Plan", "Enterprise", "Starter", "Team", "Individual"]
issues = [
"billing discrepancy of ${amount} on invoice #{inv}",
"cannot access feature '{feature}' despite paying for it",
"data export failing with error code {code}",
"SSO integration with {provider} not working",
"API rate limits hitting at {rate}/min instead of promised {expected}/min",
"webhook deliveries delayed by {hours} hours",
"user {user} locked out after password reset",
"mobile app crashing on {device} with iOS {version}",
"search returning wrong results for query '{query}'",
"file upload stuck at {percent}% for files over {size}MB",
"notification emails going to spam for domain {domain}",
"timezone showing {wrong_tz} instead of {correct_tz}",
"dashboard metrics {days} days out of date",
"cannot downgrade from {from_plan} to {to_plan}",
"GDPR data deletion request not completing for user {user_id}",
]
severities = ["critical", "high", "medium", "low"]
tickets = []
for i in range(num_tickets):
# Each ticket is genuinely unique
issue_template = issues[i % len(issues)]
issue = issue_template.format(
amount=random.randint(50, 5000),
inv=random.randint(10000, 99999),
feature=random.choice(
["advanced analytics", "custom domains", "API access", "SSO", "audit logs"]
),
code=f"ERR_{random.randint(1000, 9999)}",
provider=random.choice(["Okta", "Azure AD", "Google Workspace", "OneLogin"]),
rate=random.randint(100, 500),
expected=random.randint(1000, 5000),
hours=random.randint(1, 48),
user=f"user_{random.randint(1000, 9999)}@company{random.randint(1, 100)}.com",
device=random.choice(["iPhone 15", "iPhone 14", "iPad Pro", "iPhone 13"]),
version=random.choice(["17.2", "17.1", "16.5", "16.4"]),
query=random.choice(
["quarterly report", "user metrics", "revenue data", "team performance"]
),
percent=random.randint(45, 95),
size=random.randint(10, 500),
domain=f"company{random.randint(1, 500)}.com",
wrong_tz=random.choice(["UTC", "PST", "EST"]),
correct_tz=random.choice(["CET", "JST", "IST"]),
days=random.randint(2, 14),
from_plan=random.choice(["Enterprise", "Pro"]),
to_plan=random.choice(["Starter", "Team"]),
user_id=f"usr_{hashlib.md5(str(i).encode()).hexdigest()[:8]}",
)
tickets.append(
{
"ticket_id": f"TKT-{20000 + i}",
"customer": {
"id": f"cust_{hashlib.md5(f'customer{i}'.encode()).hexdigest()[:8]}",
"name": f"Customer {i + 1}",
"company": f"Company {chr(65 + (i % 26))}{i // 26 + 1} Inc.",
"plan": random.choice(products),
"mrr": random.randint(99, 9999),
"account_age_days": random.randint(30, 1500),
},
"issue": issue,
"severity": random.choice(severities),
"created_at": f"2024-01-{random.randint(10, 17):02d}T{random.randint(0, 23):02d}:{random.randint(0, 59):02d}:00Z",
"last_response": f"2024-01-{random.randint(15, 17):02d}T{random.randint(0, 23):02d}:{random.randint(0, 59):02d}:00Z",
"response_count": random.randint(1, 8),
"tags": random.sample(
["billing", "technical", "feature-request", "bug", "urgent", "escalated"],
k=random.randint(1, 3),
),
"assignee": None, # Unassigned - needs triage
}
)
return {
"tool": "support_queue",
"result": {"queue": "unassigned", "total_tickets": num_tickets, "tickets": tickets},
}
def generate_unique_error_traces(num_traces: int = 30) -> dict:
"""
Unique stack traces where each error is different.
This is hard for Headroom because:
- Each stack trace has different functions, line numbers
- Each error message is unique
- All errors need investigation
"""
languages = ["python", "javascript", "go", "java"]
traces = []
for i in range(num_traces):
lang = random.choice(languages)
if lang == "python":
trace = generate_python_trace(i)
elif lang == "javascript":
trace = generate_js_trace(i)
elif lang == "go":
trace = generate_go_trace(i)
else:
trace = generate_java_trace(i)
traces.append(
{
"error_id": f"err_{hashlib.md5(str(i).encode()).hexdigest()[:12]}",
"timestamp": f"2024-01-17T{10 + (i % 12):02d}:{(i * 7) % 60:02d}:00Z",
"service": random.choice(["api", "worker", "scheduler", "gateway"]),
"environment": "production",
"language": lang,
"error_type": trace["error_type"],
"message": trace["message"],
"stack_trace": trace["stack"],
"context": {
"user_id": f"user_{random.randint(10000, 99999)}",
"request_id": hashlib.md5(f"req{i}".encode()).hexdigest()[:16],
"endpoint": trace.get("endpoint", "/api/unknown"),
},
"occurrence_count": random.randint(1, 5), # Low count - each is unique
}
)
return {
"tool": "error_tracker",
"result": {"time_range": "last_24h", "total_unique_errors": num_traces, "errors": traces},
}
def generate_python_trace(seed: int) -> dict:
"""Generate a unique Python stack trace."""
error_types = [
("ValueError", f"Invalid value for parameter 'config_{seed}': expected int, got str"),
("KeyError", f"'{random.choice(['user', 'account', 'session', 'token'])}_{seed}'"),
("TypeError", f"unsupported operand type(s) for +: 'NoneType' and 'str' in field_{seed}"),
("AttributeError", f"'NoneType' object has no attribute 'process_{seed}'"),
("RuntimeError", f"Maximum recursion depth exceeded in handler_{seed}"),
("ConnectionError", f"Connection refused to service_{seed}:8080"),
(
"TimeoutError",
f"Operation timed out after {random.randint(30, 120)}s waiting for resource_{seed}",
),
]
error_type, message = random.choice(error_types)
functions = [
f"process_request_{seed}",
f"validate_input_{seed % 10}",
f"transform_data_{seed}",
f"save_to_db_{seed % 5}",
f"send_notification_{seed}",
]
stack_lines = []
for j, func in enumerate(random.sample(functions, k=random.randint(3, 5))):
line_no = random.randint(50, 500)
file_path = f"/app/services/module_{seed % 20}/{func.split('_')[0]}.py"
stack_lines.append(f' File "{file_path}", line {line_no}, in {func}')
stack_lines.append(f" result = self.handler_{j}(data)")
return {
"error_type": error_type,
"message": message,
"stack": "\n".join(stack_lines),
"endpoint": f"/api/v{random.randint(1, 3)}/{random.choice(['users', 'orders', 'products'])}/{seed}",
}
def generate_js_trace(seed: int) -> dict:
"""Generate a unique JavaScript stack trace."""
error_types = [
(
"TypeError",
f"Cannot read property '{random.choice(['map', 'filter', 'length', 'data'])}' of undefined",
),
("ReferenceError", f"config_{seed} is not defined"),
("SyntaxError", f"Unexpected token in JSON at position {random.randint(100, 1000)}"),
("RangeError", f"Maximum call stack size exceeded in recursive_{seed}"),
]
error_type, message = random.choice(error_types)
stack = f""" at processData_{seed} (/app/src/handlers/processor_{seed % 10}.js:{random.randint(50, 200)}:15)
at async handleRequest_{seed} (/app/src/routes/api_{seed % 5}.js:{random.randint(20, 100)}:23)
at async Router.dispatch (/app/node_modules/express/router.js:142:12)
at async Layer.handle (/app/node_modules/express/layer.js:95:5)"""
return {
"error_type": error_type,
"message": message,
"stack": stack,
"endpoint": f"/api/{random.choice(['graphql', 'rest', 'webhook'])}/{seed}",
}
def generate_go_trace(seed: int) -> dict:
"""Generate a unique Go stack trace."""
error_types = [
("panic", f"runtime error: index out of range [{seed}] with length {seed - 1}"),
("panic", "runtime error: invalid memory address or nil pointer dereference"),
("error", f"context deadline exceeded after {random.randint(5, 30)}s"),
("error", f"connection refused to database_{seed % 3}:5432"),
]
error_type, message = random.choice(error_types)
stack = f"""goroutine {random.randint(1, 100)} [running]:
main.processHandler_{seed}(0xc0001{seed:04x}, 0x{random.randint(1000, 9999):x})
/app/internal/handlers/handler_{seed % 10}.go:{random.randint(50, 200)} +0x{random.randint(100, 999):x}
main.(*Server).ServeHTTP_{seed}(0xc000{seed:04x}, 0x7f{random.randint(1000, 9999):x})
/app/internal/server/server.go:{random.randint(80, 150)} +0x{random.randint(100, 500):x}"""
return {
"error_type": error_type,
"message": message,
"stack": stack,
}
def generate_java_trace(seed: int) -> dict:
"""Generate a unique Java stack trace."""
error_types = [
("NullPointerException", f"Cannot invoke method on null object in Service_{seed}"),
("IllegalArgumentException", f"Parameter 'id_{seed}' cannot be negative"),
("SQLException", f"Connection to database_{seed % 3} timed out"),
("OutOfMemoryError", f"Java heap space exhausted processing batch_{seed}"),
]
error_type, message = random.choice(error_types)
stack = f"""java.lang.{error_type}: {message}
at com.app.services.Handler{seed}.process(Handler{seed}.java:{random.randint(50, 200)})
at com.app.controllers.Api{seed % 10}Controller.handle(Api{seed % 10}Controller.java:{random.randint(30, 100)})
at org.springframework.web.servlet.FrameworkServlet.service(FrameworkServlet.java:897)
at javax.servlet.http.HttpServlet.service(HttpServlet.java:750)"""
return {
"error_type": error_type,
"message": message,
"stack": stack,
}
def generate_medical_records(num_patients: int = 25) -> dict:
"""
Medical records where EVERY detail matters.
This is hard for Headroom because:
- Similar symptoms can have different diagnoses
- Missing any detail could be dangerous
- "Repetitive" info (vitals) is actually critical data
"""
conditions = [
"Type 2 Diabetes",
"Hypertension",
"Asthma",
"GERD",
"Anxiety Disorder",
"Hypothyroidism",
"Chronic Back Pain",
"Migraine",
"Allergic Rhinitis",
"Depression",
]
medications = [
"Metformin 500mg",
"Lisinopril 10mg",
"Omeprazole 20mg",
"Albuterol inhaler",
"Sertraline 50mg",
"Levothyroxine 50mcg",
"Ibuprofen 400mg PRN",
"Sumatriptan 50mg PRN",
"Loratadine 10mg",
]
records = []
for i in range(num_patients):
# Each patient has a unique combination of conditions, meds, vitals
patient_conditions = random.sample(conditions, k=random.randint(1, 4))
patient_meds = random.sample(medications, k=random.randint(1, 5))
# Vitals - these look "similar" but each patient's baseline is different
systolic = random.randint(110, 160)
diastolic = random.randint(70, 100)
records.append(
{
"patient_id": f"PT-{100000 + i}",
"name": f"Patient {chr(65 + (i % 26))}{chr(65 + ((i // 26) % 26))}",
"age": random.randint(25, 85),
"sex": random.choice(["M", "F"]),
"visit_date": f"2024-01-{random.randint(15, 17):02d}",
"chief_complaint": random.choice(
[
f"Chest pain radiating to left arm for {random.randint(1, 6)} hours",
f"Shortness of breath worsening over {random.randint(1, 14)} days",
"Severe headache, worst of life, sudden onset",
f"Abdominal pain, {random.choice(['RLQ', 'LLQ', 'epigastric'])}, {random.randint(1, 72)} hours",
f"Dizziness and {random.choice(['syncope', 'near-syncope'])} today",
f"Fever {random.randint(100, 104)}°F for {random.randint(1, 5)} days",
"Medication refill - stable on current regimen",
f"Follow-up for recent {random.choice(['hospitalization', 'procedure', 'diagnosis'])}",
]
),
"vitals": {
"bp": f"{systolic}/{diastolic}",
"hr": random.randint(60, 110),
"temp": round(random.uniform(97.5, 100.5), 1),
"resp": random.randint(12, 22),
"spo2": random.randint(94, 100),
},
"conditions": patient_conditions,
"medications": patient_meds,
"allergies": random.sample(
["Penicillin", "Sulfa", "NSAIDs", "Latex", "None"], k=random.randint(1, 2)
),
"notes": f"Patient presents with {random.choice(['acute', 'chronic', 'worsening', 'stable'])} symptoms. "
f"Last seen {random.randint(1, 12)} months ago. "
f"Compliance with medications: {random.choice(['good', 'fair', 'poor'])}. "
f"Social history: {random.choice(['non-smoker', 'former smoker', 'current smoker'])}, "
f"{random.choice(['no alcohol', 'occasional alcohol', 'daily alcohol'])}.",
}
)
return {
"tool": "ehr_query",
"result": {
"query": "today's patients",
"total_patients": num_patients,
"patients": records,
},
}
def generate_legal_discovery_docs(num_docs: int = 40) -> dict:
"""
Legal discovery documents where EVERY document must be reviewed.
This is hard for Headroom because:
- Can't skip any document - legal requirement
- "Similar" emails might have crucial differences
- Need exact quotes, not summaries
"""
senders = [f"person{i}@company.com" for i in range(1, 20)]
subjects = [
"Re: Q4 projections discussion",
"Fw: Board meeting notes",
"Re: Re: Customer complaint handling",
"Meeting tomorrow",
"Urgent: Need your input",
"Re: Project timeline update",
"Fw: Legal review needed",
"Re: Re: Re: Budget approval",
"Quick question",
"Following up",
]
docs = []
for i in range(num_docs):
sender = random.choice(senders)
recipient = random.choice([s for s in senders if s != sender])
# Each email has unique content that could be relevant
body_templates = [
f"As we discussed in the meeting on {random.randint(1, 28)}/{random.randint(1, 12)}, the numbers for Q{random.randint(1, 4)} show {random.choice(['concerning', 'promising', 'unexpected'])} trends. I think we should {random.choice(['proceed', 'hold off', 'reconsider'])} with the {random.choice(['merger', 'acquisition', 'expansion', 'restructuring'])} plan.",
f"I'm forwarding this because I think you should be aware. The customer in region {random.choice(['APAC', 'EMEA', 'Americas'])} has raised {random.choice(['serious', 'minor', 'recurring'])} concerns about our {random.choice(['pricing', 'service', 'product quality'])}. Can we discuss {random.choice(['today', 'tomorrow', 'this week'])}?",
f"Following up on your question - the {random.choice(['contract', 'agreement', 'terms'])} with {random.choice(['Vendor A', 'Vendor B', 'the client'])} does {random.choice(['', 'not '])}allow for {random.choice(['early termination', 'price adjustment', 'scope changes'])}. See clause {random.randint(1, 20)}.{random.randint(1, 9)}.",
f"Quick update: the {random.choice(['audit', 'review', 'investigation'])} team found {random.choice(['no issues', 'minor discrepancies', 'significant concerns'])} in the {random.choice(['financial records', 'compliance documents', 'HR files'])} for {random.choice(['Q1', 'Q2', 'Q3', 'Q4'])} {random.randint(2021, 2023)}.",
f"I need to flag something - the {random.choice(['employee', 'manager', 'director'])} in {random.choice(['sales', 'marketing', 'engineering'])} mentioned that {random.choice(['deadlines were missed', 'budgets were exceeded', 'protocols were bypassed'])}. Not sure if this is relevant to the case but wanted you to know.",
]
docs.append(
{
"doc_id": f"DOC-{30000 + i}",
"type": "email",
"date": f"2023-{random.randint(1, 12):02d}-{random.randint(1, 28):02d}T{random.randint(8, 18):02d}:{random.randint(0, 59):02d}:00Z",
"from": sender,
"to": [recipient],
"cc": random.sample(senders, k=random.randint(0, 3)),
"subject": random.choice(subjects),
"body": random.choice(body_templates),
"attachments": [
f"document_{random.randint(1, 100)}.{random.choice(['pdf', 'xlsx', 'docx'])}"
]
if random.random() > 0.6
else [],
"flags": random.sample(
["privileged", "responsive", "hot", "needs_review"], k=random.randint(0, 2)
),
"reviewed": False,
}
)
return {
"tool": "discovery_search",
"result": {"case": "Matter 2024-CV-1234", "total_documents": num_docs, "documents": docs},
}
# =============================================================================
# WORST-CASE SCENARIOS
# =============================================================================
@dataclass
class WorstCaseScenario:
"""A scenario where Headroom might struggle."""
name: str
description: str
why_hard: str
system_prompt: str
user_query: str
tools: list[dict]
validation_questions: list[str] # Specific questions to test recall
def create_support_triage_scenario() -> WorstCaseScenario:
"""
Support queue where every ticket is unique and important.
"""
return WorstCaseScenario(
name="Support Ticket Triage",
description="Triage 50 unique customer support tickets",
why_hard="Every ticket is unique - no patterns to compress. Each customer's problem is different. Missing any ticket means a customer gets ignored.",
system_prompt="""You are a support team lead triaging tickets.
Every ticket represents a real customer with a real problem.
You must acknowledge ALL tickets and prioritize them appropriately.
Do not skip or summarize away any customer's issue.""",
user_query="Please review all tickets in the queue and give me a prioritized action plan. I need to know about EVERY ticket - which ones need immediate attention, which can wait, and which need escalation.",
tools=[
generate_unique_support_tickets(num_tickets=50),
],
validation_questions=[
"How many critical severity tickets are there?",
"Which Enterprise customers have open tickets?",
"List all tickets related to billing issues",
"Which tickets mention SSO or authentication problems?",
],
)
def create_error_investigation_scenario() -> WorstCaseScenario:
"""
Unique errors where each needs individual investigation.
"""
return WorstCaseScenario(
name="Production Error Investigation",
description="Investigate 30 unique production errors",
why_hard="Each error has a different stack trace, different service, different root cause. Can't group them - each needs individual attention.",
system_prompt="""You are an on-call engineer investigating production errors.
Each error is unique and may indicate a different underlying issue.
Do not group or summarize - each error needs specific investigation.""",
user_query="Review all errors from the last 24 hours. For EACH error, tell me: what service, what type, and what you think the root cause might be. Don't group them - I need to know about each one individually.",
tools=[
generate_unique_error_traces(num_traces=30),
],
validation_questions=[
"How many Python errors vs JavaScript errors?",
"Which services have the most errors?",
"List all NullPointerException or nil pointer errors",
"Which errors are related to database connections?",
],
)
def create_medical_review_scenario() -> WorstCaseScenario:
"""
Medical records where every detail matters.
"""
return WorstCaseScenario(
name="Medical Record Review",
description="Review 25 patients for today's clinic",
why_hard="Every patient's vitals, conditions, and medications are unique. 'Similar' symptoms could mean very different things. Can't summarize - details save lives.",
system_prompt="""You are a physician reviewing today's patient list.
Every patient's details matter - similar symptoms may need different treatment.
Pay attention to vital signs, medication lists, and allergies.
Never assume two patients with similar complaints have the same issue.""",
user_query="Review all patients on today's schedule. Flag any concerning vitals, potential drug interactions, or high-acuity complaints. Give me a brief on EACH patient.",
tools=[
generate_medical_records(num_patients=25),
],
validation_questions=[
"Which patients have BP over 140 systolic?",
"Which patients are on Metformin?",
"List patients with chest pain or cardiac symptoms",
"Which patients have drug allergies we should note?",
],
)
def create_legal_discovery_scenario() -> WorstCaseScenario:
"""
Legal documents where completeness is required.
"""
return WorstCaseScenario(
name="Legal Discovery Review",
description="Review 40 documents for legal discovery",
why_hard="Legal requirement to review EVERY document. Similar-looking emails may have crucial differences. Need exact recall - summaries aren't acceptable in court.",
system_prompt="""You are a legal assistant reviewing discovery documents.
EVERY document must be accounted for - missing one could be sanctions.
Pay attention to dates, senders, and specific language used.
Similar documents may have legally significant differences.""",
user_query="Review all documents and categorize them. For each document, note: the date, sender, key topics, and whether it seems relevant to the case. I need a complete accounting.",
tools=[
generate_legal_discovery_docs(num_docs=40),
],
validation_questions=[
"How many documents mention 'audit' or 'investigation'?",
"List all documents with attachments",
"Which documents are flagged as 'privileged'?",
"How many documents were sent in Q4 2023?",
],
)
# =============================================================================
# BENCHMARK RUNNER
# =============================================================================
@dataclass
class BenchmarkResult:
"""Result from running a scenario."""
scenario_name: str
mode: str
input_tokens: int
output_tokens: int
cost_usd: float
latency_ms: float
answer: str
validation_scores: dict # Scores for each validation question
def count_tokens(text: str) -> int:
"""Simple token estimation."""
return len(text) // 4
def validate_answer(answer: str, scenario: WorstCaseScenario) -> dict:
"""
Check if the answer addresses all validation questions.
Returns dict of question -> (found keywords, score).
"""
scores = {}
answer_lower = answer.lower()
for question in scenario.validation_questions:
# Extract key terms from question
key_terms = [w for w in question.lower().split() if len(w) > 4]
found = sum(1 for term in key_terms if term in answer_lower)
score = found / len(key_terms) if key_terms else 0
scores[question] = {
"terms_found": found,
"terms_total": len(key_terms),
"score": round(score, 2),
}
return scores
def run_scenario(
client: Any, scenario: WorstCaseScenario, mode: str, model: str = "gpt-4o-mini"
) -> BenchmarkResult:
"""Run a single scenario."""
messages = [
{"role": "system", "content": scenario.system_prompt},
{"role": "user", "content": scenario.user_query},
]
# Add tool results with proper format
for tool_output in scenario.tools:
tool_call_id = f"call_{hashlib.md5(tool_output['tool'].encode()).hexdigest()[:8]}"
messages.append(
{
"role": "assistant",
"content": None,
"tool_calls": [
{
"id": tool_call_id,
"type": "function",
"function": {"name": tool_output["tool"], "arguments": "{}"},
}
],
}
)
messages.append(
{
"role": "tool",
"tool_call_id": tool_call_id,
"content": json.dumps(tool_output["result"], indent=2),
}
)
messages.append({"role": "user", "content": "Please provide your complete analysis now."})
start = time.time()
try:
response = client.chat.completions.create(
model=model,
messages=messages,
max_tokens=4000, # Allow longer responses
)
latency = (time.time() - start) * 1000
answer = response.choices[0].message.content
input_tokens = response.usage.prompt_tokens
output_tokens = response.usage.completion_tokens
# GPT-4o-mini pricing
cost = (input_tokens * 0.00015 + output_tokens * 0.0006) / 1000
validation_scores = validate_answer(answer, scenario)
except Exception as e:
print(f" Error: {e}")
return BenchmarkResult(
scenario_name=scenario.name,
mode=mode,
input_tokens=count_tokens(json.dumps(messages)),
output_tokens=0,
cost_usd=0,
latency_ms=0,
answer=f"Error: {e}",
validation_scores={},
)
return BenchmarkResult(
scenario_name=scenario.name,
mode=mode,
input_tokens=input_tokens,
output_tokens=output_tokens,
cost_usd=cost,
latency_ms=latency,
answer=answer,
validation_scores=validation_scores,
)
def run_worst_case_benchmark(api_key: str = None) -> dict:
"""Run the complete worst-case benchmark."""
if api_key is None:
api_key = os.environ.get("OPENAI_API_KEY")
if not api_key:
raise ValueError("OPENAI_API_KEY required")
print("=" * 70)
print("HEADROOM WORST-CASE BENCHMARK")
print("Testing scenarios where compression may hurt performance")
print("=" * 70)
# Create clients
import tempfile
from openai import OpenAI
baseline_client = OpenAI(api_key=api_key)
if HEADROOM_AVAILABLE:
db_path = os.path.join(tempfile.gettempdir(), "headroom_worst_case.db")
headroom_client = HeadroomClient(
original_client=OpenAI(api_key=api_key),
provider=OpenAIProvider(),
store_url=f"sqlite:///{db_path}",
default_mode="optimize",
)
else:
print("WARNING: Headroom not available, running baseline only")
headroom_client = None
scenarios = [
create_support_triage_scenario(),
create_error_investigation_scenario(),
create_medical_review_scenario(),
create_legal_discovery_scenario(),
]
results = []
for scenario in scenarios:
print(f"\n{'=' * 60}")
print(f"Scenario: {scenario.name}")
print(f"Description: {scenario.description}")
print(f"WHY THIS IS HARD: {scenario.why_hard}")
print("=" * 60)
# Calculate raw size
raw_size = sum(len(json.dumps(t["result"], indent=2)) for t in scenario.tools)
print(f"\nRaw tool output size: {raw_size:,} chars (~{raw_size // 4:,} tokens)")
# Run baseline
print("\n[1/2] Running BASELINE...")
baseline_result = run_scenario(baseline_client, scenario, "baseline")
print(f" Input tokens: {baseline_result.input_tokens:,}")
print(f" Output tokens: {baseline_result.output_tokens:,}")
print(f" Cost: ${baseline_result.cost_usd:.4f}")
avg_baseline_score = (
sum(v["score"] for v in baseline_result.validation_scores.values())
/ len(baseline_result.validation_scores)
if baseline_result.validation_scores
else 0
)
print(f" Validation score: {avg_baseline_score:.1%}")
results.append(baseline_result)
# Run Headroom
if headroom_client:
print("\n[2/2] Running HEADROOM...")
headroom_result = run_scenario(headroom_client, scenario, "headroom")
print(f" Input tokens: {headroom_result.input_tokens:,}")
print(f" Output tokens: {headroom_result.output_tokens:,}")
print(f" Cost: ${headroom_result.cost_usd:.4f}")
avg_headroom_score = (
sum(v["score"] for v in headroom_result.validation_scores.values())
/ len(headroom_result.validation_scores)
if headroom_result.validation_scores
else 0
)
print(f" Validation score: {avg_headroom_score:.1%}")
results.append(headroom_result)
# Compare
if baseline_result.input_tokens > 0:
token_change = (
headroom_result.input_tokens - baseline_result.input_tokens
) / baseline_result.input_tokens
quality_change = avg_headroom_score - avg_baseline_score
print("\n 📊 COMPARISON:")
print(
f" Token change: {token_change:+.1%} ({'saved' if token_change < 0 else 'INCREASED'})"
)
print(
f" Quality change: {quality_change:+.1%} ({'preserved' if quality_change >= -0.1 else 'DEGRADED'})"
)
if quality_change < -0.1:
print(" ⚠️ WARNING: Quality degraded significantly!")
# Summary
print("\n" + "=" * 70)
print("WORST-CASE BENCHMARK SUMMARY")
print("=" * 70)
baseline_results = [r for r in results if r.mode == "baseline"]
headroom_results = [r for r in results if r.mode == "headroom"]
print(f"\n{'Scenario':<30} {'Baseline Tokens':>15} {'Headroom Tokens':>15} {'Quality Δ':>12}")
print("-" * 72)
for br in baseline_results:
hr = next((r for r in headroom_results if r.scenario_name == br.scenario_name), None)
if hr:
b_score = (
sum(v["score"] for v in br.validation_scores.values()) / len(br.validation_scores)
if br.validation_scores
else 0
)
h_score = (
sum(v["score"] for v in hr.validation_scores.values()) / len(hr.validation_scores)
if hr.validation_scores
else 0
)
quality_delta = h_score - b_score
print(
f"{br.scenario_name:<30} {br.input_tokens:>15,} {hr.input_tokens:>15,} {quality_delta:>+11.1%}"
)
return {
"baseline": [
{
"scenario": r.scenario_name,
"tokens": r.input_tokens,
"cost": r.cost_usd,
"validation": r.validation_scores,
}
for r in baseline_results
],
"headroom": [
{
"scenario": r.scenario_name,
"tokens": r.input_tokens,
"cost": r.cost_usd,
"validation": r.validation_scores,
}
for r in headroom_results
],
}
if __name__ == "__main__":
results = run_worst_case_benchmark()
with open("worst_case_benchmark_results.json", "w") as f:
json.dump(results, f, indent=2)
print("\nResults saved to worst_case_benchmark_results.json")
|