From Self-Assessment to Evidence: An AI-Driven Validation Framework for Business Impact Analysis in Telecommunications

Business impact analysis (BIA) underpins every business continuity programme, yet in most organisations it is still assembled from self-reported values that are rarely tested before a real disruption. At the national telecom operator that motivated this work, the annual BIA campaign takes about 16 weeks, a month of which is rework, and almost every check is made by email or telephone. This paper presents Omni-BIA, a framework that treats each declared recovery time objective (RTO), recovery point objective (RPO) and dependency as a claim to be tested against evidence the organisation already holds, in line with ISO 22301:2019 and ISO/TS 22317:2021. Deterministic rules check values against systems of record. Machine learning (ML) predicts recovery times, flags outliers and mines undeclared dependencies, and generative artificial intelligence (AI) reads service level agreements (SLAs). A Monte Carlo digital twin then tests whole recovery chains. The framework was demonstrated with reproducible code on a synthetic telecom operator with 30 services, 105 processes and 212 known issues. Its checks raised 178 flags with 98.9% precision and 83.0% recall; recall was 95.1% over the fault types they target. In a simulated data-centre outage, a priority 2 service whose dependencies each have a median recovery time inside its 24-hour RTO still had only a 10.2% chance of meeting that RTO. The dependencies recover one after another: the median was 28.4 hours. The twin's drill-based ranges contained only 22 of 40 past incidents; after calibration on incident data, coverage averaged 82.5% against a target of 80%. Manual touchpoints fell by an estimated 62% to 83%. Because the data are synthetic, the results show what the approach can detect, not how often real BIAs fail. For practitioners, the work shows how a BIA can move from self-assessment to evidence while people remain accountable for every decision.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-13
DOI
https://doi.org/10.5281/zenodo.24202806
Primary Topic
Supply Chain Resilience and Risk Management
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

From Self-Assessment to Evidence: An AI-Driven Validation Framework for Business Impact Analysis in Telecommunications

Muhammad Zeeshan Khan, Muhammad Affan Hashmi
Zenodo (CERN European Organization for Nuclear Research)
Supply Chain Resilience and Risk Management
preprint

From Self-Assessment to Evidence: An AI-Driven Validation Framework for Business Impact Analysis in Telecommunications

Muhammad Zeeshan Khan, Muhammad Affan Hashmi
preprint en

Abstract

Business impact analysis (BIA) underpins every business continuity programme, yet in most organisations it is still assembled from self-reported values that are rarely tested before a real disruption. At the national telecom operator that motivated this work, the annual BIA campaign takes about 16 weeks, a month of which is rework, and almost every check is made by email or telephone. This paper presents Omni-BIA, a framework that treats each declared recovery time objective (RTO), recovery point objective (RPO) and dependency as a claim to be tested against evidence the organisation already holds, in line with ISO 22301:2019 and ISO/TS 22317:2021. Deterministic rules check values against systems of record. Machine learning (ML) predicts recovery times, flags outliers and mines undeclared dependencies, and generative artificial intelligence (AI) reads service level agreements (SLAs). A Monte Carlo digital twin then tests whole recovery chains. The framework was demonstrated with reproducible code on a synthetic telecom operator with 30 services, 105 processes and 212 known issues. Its checks raised 178 flags with 98.9% precision and 83.0% recall; recall was 95.1% over the fault types they target. In a simulated data-centre outage, a priority 2 service whose dependencies each have a median recovery time inside its 24-hour RTO still had only a 10.2% chance of meeting that RTO. The dependencies recover one after another: the median was 28.4 hours. The twin's drill-based ranges contained only 22 of 40 past incidents; after calibration on incident data, coverage averaged 82.5% against a target of 80%. Manual touchpoints fell by an estimated 62% to 83%. Because the data are synthetic, the results show what the approach can detect, not how often real BIAs fail. For practitioners, the work shows how a BIA can move from self-assessment to evidence while people remain accountable for every decision.

Zenodo (CERN European Organization for Nuclear Research)
Industry, innovation and infrastructure
Supply Chain Resilience and Risk Management
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon