Run Date: 2026-10-24 01:46:20.102563 Copyright (c) Citizens' Oversight
This report provides a central location for all reports produced by AuditEngine. It lists every report of the pipeline from the start. The link of a report opens once its stage has run.
The table below lists the attributes of this election. If the audit was run in cooperative workflow prior to any statement of the "Total Ballots Cast" then it will be missing below until that value is added to the job settings file.
| Election Name | San Francisco, California, 2024 General Election |
| District Population | 826079 |
| District Population Year | 2025 |
| Registered Voters | 522265 |
| Total ballots cast in district (official report) | 412231 |
| Ballots cast in profiled scope (official report) | 34541 |
| Voter Turnout of Registered Voters | 78.93% |
| Voter Turnout of Population | 49.9% |
| District Partisan Bias | |
| Voting Equipment Vendor | Dominion |
| Job folder (required) | s3://us-east-1-audit-engine-jobs/US/CA/CA_SanFrancisco_20241105_pilot_d05/ |
| Workflow methodology | public_oversight |
| Archives folder (required) | s3://us-east-1-audit-engine-election-data/US/CA/CA_SanFrancisco_20241105/ |
| Number of LAT Images Archives spec'd | 0 |
| Number of LAT Results Files spec'd | 0 |
| EIF File Specified (required) | EIF_CA_SanFrancisco_20241105_pilot_d05.csv |
| Number of style_masters archives spec'd | 0 |
| Number of Ballot Archives spec'd | 0 |
| Number of CVR files | 11 |
| Number of Verification Archives spec'd | 0 |
| Narrative Report | Narrative Report (Not Available) |
| Ballot Viewer App | BallotViewer (Not Available) |
| AdjudiTally App | AdjudiTally (Not Available) |
The following automatically generated reports are the most important reports for casual reviewers of the audit results.
| Stage / Report | Report Description |
|---|---|
| gen_final_report Final Report |
This report states the result of the audit. It is not written yet. |
| gen_source_audit_report source_Totals_Report |
This report provides a review of each contest in detail based on the ballot images in the 'source' archives. If the CVR is available when this report is created, then the totals from the CVR will be included with the difference provided. |
| gen_cmpcvr_report Reconciliation Report |
This report provides details of disagreements between the voting system and the auditing result, as well as variants, such as write-ins, overvotes, and ballots which are 'gray-flagged' by AuditEngine for potential further review. This version of this report includes comparisons with the CVR, any manual results, and any adjudications. |
The reports listed here are in the same order as the pipeline stages that were executed to produce these audit results.
Many of these reports exist to allow for auditing analysts to check the progress of the audit.
All other supporting files and intermediate files can be found in the "Pipeline Report".
Most users may find what they need under "Main Reports", above.
| Stage / Report | Report Description |
|---|---|
| precheck precheck Report |
|
| gen_master_report_index Master Report Index |
This report lists all reports of the audit. It was called the Final Report. It is rebuilt whenever the pipeline definition changes. |
| verify_source_hashes Source Hashes Report |
This report checks the election files against the hashes the county published. A matching hash shows the file is the one the county published. Each hash is kept in a sidecar file with the S3 ETag of the file it was made from. A later run reuses that hash while the ETag is unchanged, and says when it was made. |
| gen_biabif Ballot Image Archive Metadata Report |
This report details information about the ballot image archives provided by the election district in terms of the metadata regarding each image file and its size, as well as the number of images in each file, repeated image files, etc. |
| check_crc_duplicates CRC Possible Duplicates Report |
This report lists ballot images whose page CRC32 checksums match another image in the archive. Each match is a possible duplicate scan. Dominion only. |
| parse_eif Election Information File (EIF) Parse Report |
The Election Information File (EIF) contains the full description of the election in terms of what contests are included and what options are listed for each contest. The EIF file is usually generated from the CVR, if it is available (Public Oversight Workflow) or from other sources, such as the Logic and Accuracy Test results, in Cooperative Workflow. The EIF file can also be edited by hand, and commonly the number of write-ins need to be modified. |
| gen_cvrbif CVR BIF Report: CVR integrity checks |
This report checks the finished CVR tables against each other and against the source CVR files. It finds repeated ballot rows, missing or extra ballots, and totals that do not add up. It is a report only. A FAIL does not stop the pipeline. Dominion JSON CVRs only. |
| gen_biacvr_report BIA-CVR Join Report: images and CVR records without a partner |
This report lists ballot images with no CVR record, and CVR records with no image, after the two are joined. It runs before the full image scan, so a missing archive or an unreadable ballot id is found early. It also lists election folder files the job does not read. It is a report only. |
| imagematch Image_Match_Report |
This report is the result of a complete examination of all ballot images to create a cryptographic hash of each image file, to make it easy to local duplicated image files. There will be a hash value for each side of each sheet, but sometimes no "back" image is provided, so these may not be the same. When repeated images are detected, the first images detected are kept and all repeats of those are "skipped". If repeated images are detected, then it could have been caused by uploading results from the same machine multiple times. |
| gen_acre_report ACRE Report |
This report launches the ACRE (All Contests Review & Evaluation) ballot-by-ballot review: a bounded sample of ballots, drawn to cover every contest option, undervote, overvote, and write-in-used check across every ballot style, grouped into review-priority rounds. Open the ACRE Review App from here to compare each selected ballot's image against its CVR record and record findings. |
| gen_acre_report ACRE Statistics Report |
This report documents the ACRE sample's run parameters, methodology, and coverage in detail: how many ballots were selected and why, the review-tier/round breakdown, oval and strata coverage progress, and per-contest margin, RLA sample size, and coverage statistics. Use this report to verify the sample itself was sound, not to review individual ballots. |
| gen_acre_report ACRE Review App (link for reviewers) |
Opens the standalone ACRE Review App with this run's ballot sample already loaded: compare each selected ballot's image against its CVR record, filter by round or contest, mark each contest as matching or not, and export the review. |
| gen_bif_report bif: Combined Metadata Report |
This report is similar to the "nocvr_bif: Combined Metadata Report" but now the CVR information has been combined and correlated. |
| writein_inspection_report Write-in inspection report |
This report shows write-in ballot images next to their OCR'd text, for manually reviewing whether the audit's write-in reading is accurate. |
| gen_template_tasklists Generate Templates Tasklist Report |
This report lists the style/page tasks queued for template generation, before the (often long-running) gentemplates stage actually runs them. |
| gen_templates_report Templates Report |
This report shows the generated ballot style templates -- the target/oval positions AuditEngine will use to read votes -- for visual review before extraction. |
| register_job_with_frontend Run TargetMapper App |
Link to launch the TargetMapper app for this job, for manually reviewing and promoting AI/OCR- proposed option-name mappings. |
| gen_option_proofs_report Option Proofs Report |
This report is to allow human-eye review of each option on all styles. We find that human-perception is particularly adept at finding any deviations among many that are the same. |
| gen_styles_report Style Redline Proofs Report |
This report is also called the "Redline Proofs report" and it provides an image of every style with the oval outlined with a rectangle and then marked with the name of the contest and options of that contest. We have found this report is relatively hard for human perception to review, and thus we have also the Option Proofs Report, which is much easier to check that each option is correctly mapped. However, it may be the case that an oval is left out, and it would not be shown at all on the option proofs report. Therefore, we still need this report, which should be reviewed only to verify that all ovals have been redlined. |
| bmd_map_report BMD map report |
This report shows how each BMD (Ballot Marking Device) summary page's printed contest and option text was matched to the EIF, for review before extraction relies on that mapping. |
| gen_extraction_coverage_report Extraction coverage report |
This report checks that every ballot sheet sent to vote extraction has a result or an exception. It also shows how many sheets the style scope leaves out, and whether every chunk was built. |
| gen_source_audit_report source_Totals_Report |
This report provides a review of each contest in detail based on the ballot images in the 'source' archives. If the CVR is available when this report is created, then the totals from the CVR will be included with the difference provided. |
| parse_results_stdresults results Results Parsing Report |
|
| import_adjudications_cmpcvr import_adjudications_cmpcvr Report |
This report is available if adjudications have been imported after the comparison process. Once data has been imported, an adjudicated column will appear which corrects the Audit system results. |
| gen_cmpcvr_report Reconciliation Report |
This report provides details of disagreements between the voting system and the auditing result, as well as variants, such as write-ins, overvotes, and ballots which are 'gray-flagged' by AuditEngine for potential further review. This version of this report includes comparisons with the CVR, any manual results, and any adjudications. |
| gen_final_report Final Report |
This report states the result of the audit. It is not written yet. |
| Pipeline Report Pipeline Report |
This report is updated as the processing pipeline is run, and provides the versions of all input files and outputs from each stage. |
Reports listed here were produced by manually running --op export_cmpcvr_group for a specific
discrepancy group (e.g. x_uv, uv_x, ov) -- a full per-record export of every ballot in that
group, beyond the capped number of examples shown in the main Reconciliation Report. Not every job
will have these; they are run on request during manual review, not as part of the standard
pipeline.
| Manual Export | Report Description |
|---|---|
| export_cmpcvr_group --cmpcvr_export_group bmd_x_uv 'bmd_x_uv' Reconciliation Group Export |
Unadjudicated disagreed contests with CVR undervotes (gray OK, Only BMDs). This can occur if the AuditEngine BMD OCR parsing is making an error. |
| export_cmpcvr_group --cmpcvr_export_group bmd_x_y 'bmd_x_y' Reconciliation Group Export |
Unadjudicated disagreed contests with votes (gray OK, Only BMDs) |