Skip to main content

Detection Scorecard

Detection accuracy scorecard

Epic Fill measures field-detection quality against a fixed suite of ATS-shaped HTML fixtures — Greenhouse, Workday, Lever, iCIMS, and peers — with hand-labeled canonical field types. The shipped extension classifier is bundled and scored on every labeled field. This is an internal product-quality scorecard, not a competitor comparison.

Aggregate accuracy

99.5%

182 / 183 fields

Micro-F1

99.5%

extension field-scanner

Fixtures

31

ATS-shaped pages

Labeled fields

183

hand-labeled selectors

Generated 2026-07-17 07:50:51 UTC · engine extension field-scanner

FixtureFieldsCorrectAccuracy
ADP
adp
77100.0%
Ashby
ashby
44100.0%
BambooHR
bamboohr
77100.0%
BreezyHR
breezyhr
66100.0%
ClearCompany
clearcompany
77100.0%
Dayforce
dayforce
77100.0%
Google Forms Intake
google-forms-intake
44100.0%
Greenhouse
greenhouse
5480.0%
iCIMS
icims
44100.0%
iFrame Form
iframe-form
33100.0%
JazzHR
jazzhr
66100.0%
Jobvite
jobvite
88100.0%
Lever
lever
44100.0%
Multi-Step Wizard
multi-step-wizard
77100.0%
Multilingual (DE)
multilingual-de
44100.0%
Multilingual (ES)
multilingual-es
55100.0%
Negative Landing Page
negative-landing-page
00100.0%
Oracle HCM
oracle-hcm
88100.0%
Paylocity
paylocity
66100.0%
React Controlled
react-controlled
44100.0%
Recruitee
recruitee
66100.0%
Repeated Employment
repeated-employment
1212100.0%
School District
school-district
1212100.0%
Shadow DOM
shadow-dom
33100.0%
SmartRecruiters
smartrecruiters
66100.0%
SuccessFactors
successfactors
88100.0%
Taleo
taleo
66100.0%
UKG
ukg
77100.0%
USAJobs Questionnaire
usajobs-questionnaire
44100.0%
Workable
workable
66100.0%
Workday
workday
77100.0%

Per-type F1 (top 12 by support)

Precision, recall, and F1 for the most frequently labeled canonical types in the fixture suite.

Canonical typeSupportPrecisionRecallF1
email28100.0%100.0%100.0%
phone27100.0%100.0%100.0%
first_name22100.0%100.0%100.0%
last_name21100.0%100.0%100.0%
city13100.0%100.0%100.0%
street_19100.0%100.0%100.0%
employer8100.0%100.0%100.0%
postal_code8100.0%100.0%100.0%
job_title6100.0%100.0%100.0%
cover_letter6100.0%83.3%90.9%
full_name5100.0%100.0%100.0%
start_date4100.0%100.0%100.0%

Methodology

Each fixture is a captured ATS-shaped HTML page with a sibling label manifest mapping CSS selectors to expected canonical field types. The shipped extension field classifier is esbuild-bundled into a browser IIFE and executed against every labeled field via jsdom (including shadow-root piercing where needed). A field counts as correct when the predicted canonical type matches the label.

Scoring

Aggregate accuracy is correct predictions over labeled fields. Micro-F1 is computed from a multiclass confusion matrix (true positives when predicted equals expected; otherwise a false negative for the expected type and a false positive for the predicted type). Per-type precision, recall, and F1 use the same matrix.

Reproducibility

Regenerate this scorecard from the repository root with node e2e/gold-ext/scripts/emit-detection-scorecard.mjs. Output is written to frontend/lib/detection-scorecard.json and served statically on this page.

Why Epic Fill

The user should feel the performance, not the machinery behind it.

Premium control. Simple interaction. Pricing that tracks real system effort.

Maximum sophistication. Minimum complexity. Zero visible compromise.

Quality by defaultHidden sophisticationReal system power
Encrypted at rest & in transit
Mandatory Review Wall
No auto-submission
GDPR-ready data controls

"Nothing submits until you approve it."

© 2026 Epic Fill. All rights reserved.

Built with Next.js & Tailwind · Your data stays yours