InnoAI
AI & Data Science MUSTKiongozi-Sheria AI
Problem Statement
Government institutions produce a large number of official documents as part of their daily administrative operations. These may include correspondence, notices, reports, requests, approvals, human resource documents, procurement-related documents, disciplinary documents, and other administrative documents. Although these documents are governed by laws, regulations, government guidelines, and institutional procedures, ensuring compliance during document preparation can be difficult. Users may prepare a document that contains, missing mandatory information, incorrect or incomplete procedures, inappropriate wording, incorrect application of a requirement, a required step that has not been followed, information that does not match the applicable institutional procedure. In many cases, these issues are identified through manual review by another officer or supervisor. This can increase processing time, create repeated corrections, and result in inconsistent document quality.
Proposed Solution
Kiongozi-Sheria AI is an intelligent compliance engine designed to support the preparation, review, and processing of official government documents. It acts as an intelligent compliance layer within the document workflow, checking documents against applicable laws, regulations, government procedures, and institutional requirements. The system supports two main inputs: real-time text checking while a user prepares a document, and intelligent document processing for uploaded files such as .docx and .pdf. It analyzes the document, identifies potential compliance issues, and provides the relevant evidence, legal or institutional reference, and required correction. After correction, the system re-checks the document to verify whether the identified requirements have been satisfied. If a mandatory compliance requirement remains unsatisfied, the system blocks the document from being submitted or proceeding to the next stage until the issue is resolved or referred for authorized human review. The system can be integrated with existing government document workflows where supported.
Innovation
The core innovation of Kiongozi-Sheria AI is the combination of AI-based document understanding, legal intelligence, and mandatory compliance enforcement within the government document workflow. 1. AI Understanding: The system uses Natural Language Processing to understand document content, identify meaningful statements, and determine the relevant regulatory domain. 2. Legal Intelligence: The system retrieves applicable laws, regulations, procedures, and institutional requirements from the Legal and Administrative Knowledge Base and provides the relevant supporting references. 3. Compliance Enforcement: The system checks whether the identified requirements are satisfied, guides the user in correcting detected issues, and re-checks the document. Where a defined mandatory requirement remains unresolved, the system prevents the document from proceeding to the next workflow stage or refers it for authorized human review.
Expected Impact
1. Compliance Enforcement: Improves compliance by preventing documents with unresolved mandatory requirements from proceeding through the official workflow. 2. Workflow Optimization: Reduces repetitive manual compliance checking, allowing legal and administrative reviewers to focus on complex cases that require human judgment. 3. Error Risk Mitigation: Minimizes avoidable procedural and compliance errors that may lead to document rejection, repeated correction, or administrative review. 4. Continuous Capacity Building: Improves users’ understanding of applicable legal and administrative requirements through contextual explanations, statutory references, and correction guidance.