Challenges in Adopting Artificial Intelligence Based User Input Verification Framework in Reporting Software Systems
Bibliographic record
Abstract
Artificial intelligence is driving new industrial solutions for challenging problems once considered impossible. Many large-scale companies use AI to identify opportunities to improve business processes and products. Despite the promise and perils of AI, many traditional software systems (e.g., taxation or reporting) are implemented without AI in mind. Adopting AI-based capabilities in such software can be challenging due to a lack of resources and uncertainties in requirements. This paper documents our experience working with our industry partner on adopting AI capabilities in enterprise software. The enterprise software receives and processes thousands of user inputs with different configuration settings daily, which makes manual user input verification infeasible. To assist our industry partner, we design and integrate an AI-based input verification framework into the software. However, during the design and integration of the framework, we encounter many challenges that range from the requirement engineering process to the development, adoption, and verification process. We discuss the challenges we encountered and their corresponding solutions while working with our industrial partner to integrate the AI-based input verification framework into their non-AI software. Our experience report may provide valuable insight to practitioners and researchers on better integrating AI-based capabilities with existing software systems.
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.093 | 0.120 |
| Meta-epidemiology (narrow) | 0.001 | 0.002 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.003 | 0.002 |
| Science and technology studies | 0.002 | 0.007 |
| Scholarly communication | 0.010 | 0.014 |
| Open science | 0.007 | 0.006 |
| Research integrity | 0.004 | 0.008 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".