Robotic Process Automation for the Extraction of Audit Information: A Use Case
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
SUMMARY The reconciliation of audit evidence to the audit subject matter is a key and recurring audit procedure. Before reconciling information, data needs to be extracted from the audit subject matter, which is often in a Portable Document Format (PDF). Reconciliations are a recurring task for every new version of the audit subject matter. Large audit firms typically “offshore” simple and repetitive audit tasks such as reconciliations to shared service centers. Offshoring however comes at the expense of coordination costs, delays in the process, and challenges regarding the liability risk to the auditor. This paper presents an open-source algorithm to extract data from (draft) annual reports (PDF files) using Python to automate, rather than outsource, the data extraction for reconciliations. The algorithm resulted in a significant time saving for the audit of a large Dutch asset management firm. Researchers apply the algorithm to minimize hand-collection of financial statement data. Data Availability: The algorithm this paper presents is open-source and publicly available. JEL Classifications: M42; G23; G29.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it