AI projects in INPS
| AI system | Operational period | Volume processed | Automation rate | Estimated time saved | Key performance notes |
|---|---|---|---|---|---|
| Automated routing system (CRMweb level I and II) | April 2024–September 2025 | 1.8 M tickets analysed | 507K tickets routed directly to back-office specialists | ∼25,350 h (3,168 person-days) | Bypasses front-office triage, enabling direct assignment to appropriate back-office personnel. Average manual processing: 3 min (front-office) + 7 min (back-office) per ticket |
| AI Dashboard for video guide creation | January–September 2025 | 3.4 M contact motivations received | 3.3 M analysed automatically (96.6% automation) | ∼55,000 h (6,875 person-days) | Enables proactive content creation by identifying most frequent citizen enquiries. Communications staff can develop targeted video guides and social media content, reducing incoming query volume. Real-time monitoring of contact patterns by benefit type |
| Automated PEC classification | April 2021–September 2025 | 4.8 M certified emails analysed | 3.2 M automatically routed (65.9% automation) | ∼136,000 h (17,000 person-days) | 85% precision rate. On-premise BERT-based transformer model ensures GDPR compliance and complete institutional control. Average manual processing: 3 min per email. System handles heterogeneous legal and administrative correspondence |
| INPSieme welfare program document verification | February–September 2025 | 207,000 applications analysed | Full automation of initial document verification | ∼17,250 h (2,156 person-days) | 100% precision in document completeness checking for study travel reimbursements. Estimated 50% reduction in case processing time. Average manual review: 10 min per application. Note: First implementation phase, full efficiency gains pending as staff adapt to modified procedures |
| AI system | Operational period | Volume processed | Automation rate | Estimated time saved | Key performance notes |
|---|---|---|---|---|---|
| Automated routing system (CRMweb level I and II) | April 2024–September 2025 | 1.8 M tickets analysed | 507K tickets routed directly to back-office specialists | ∼25,350 h (3,168 person-days) | Bypasses front-office triage, enabling direct assignment to appropriate back-office personnel. Average manual processing: 3 min (front-office) + 7 min (back-office) per ticket |
| AI Dashboard for video guide creation | January–September 2025 | 3.4 M contact motivations received | 3.3 M analysed automatically (96.6% automation) | ∼55,000 h (6,875 person-days) | Enables proactive content creation by identifying most frequent citizen enquiries. Communications staff can develop targeted video guides and social media content, reducing incoming query volume. Real-time monitoring of contact patterns by benefit type |
| Automated PEC classification | April 2021–September 2025 | 4.8 M certified emails analysed | 3.2 M automatically routed (65.9% automation) | ∼136,000 h (17,000 person-days) | 85% precision rate. On-premise BERT-based transformer model ensures GDPR compliance and complete institutional control. Average manual processing: 3 min per email. System handles heterogeneous legal and administrative correspondence |
| INPSieme welfare program document verification | February–September 2025 | 207,000 applications analysed | Full automation of initial document verification | ∼17,250 h (2,156 person-days) | 100% precision in document completeness checking for study travel reimbursements. Estimated 50% reduction in case processing time. Average manual review: 10 min per application. |
Note(s): PEC: Posta Elettronica Certificata, the italian certified e-mail system
Sharing content requires targeting cookies to be enabled. Please update your cookie preferences to use this feature.