ARGUS Survey of AI Driven Personal Finance Analytics and Budget Forecasting (International Journal of Scientific Research in Computer Science Engineering and Information Technology)

Summary

• The survey reviews a locked corpus of 25 papers published between 1 January 2025 and 30 September 2026 to evaluate AI evidence in personal finance.p.1
• Observational surveys represent 11 of the 25 papers (44%), while human experiments account for 6 papers (24%).p.8
• Direct personal transaction classification and budget prediction is evaluated in only 1 paper (4%), designated as P02.p.9
• The inspected ARGUS implementation passed 205 automated checks across API, mobile JS, AI core, and native notification components.p.12
• Automated code line coverage reached 80.11% for API modules, 91.26% for AI core modules, and 52.43% for mobile JavaScript.p.12

Key insights

1. Observational surveys constitute the largest single methodology group in the reviewed literature at 44% of total papers.p.8
2. Direct computational modeling of personal transactions and budget prediction is addressed by only 2 of the 25 corpus studies.p.8
3. Corporate crisis warning model P14 demonstrated an 88.2% accuracy with a 5.6% false alarm rate.p.9
4. In the ARGUS categorization engine, explicit user corrections receive a fixed confidence score of 0.99.p.11
5. ARGUS InsightGenerator triggers period-change alerts at absolute changes of 1,000 basis points with 0.88 confidence for five or more transactions.p.11
6. Evaluation of Ask ARGUS recorded acceptable grounded responses for 10 out of 10 controlled questions.p.12
7. Fresh-process launch latency across five runs averaged 1,586.6 ms with a median of 1,662 ms.p.12