AI-Assisted Scenario Generation for Faster Cybersecurity Consulting
Challenge
A cybersecurity consulting team based in the United Kingdom built their client engagements around diagnostic scenarios — structured simulations designed to assess organizational readiness for real-world threats. These scenarios were central to their service offering, but the process of creating them was entirely manual and time-intensive.
A single scenario could take up to two weeks to produce. Consultants would begin with drafting, then move through multiple rounds of client review, incorporate feedback, refine the content, and repeat the cycle until the scenario met the required standard. The iterative review process consumed a significant share of each engagement’s timeline — time that could otherwise be spent on actual analysis and advisory work.
As demand for their services grew, the manual content creation bottleneck became a direct constraint on the team’s capacity to take on new clients and deliver engagements efficiently. The firm needed a way to dramatically accelerate scenario production without sacrificing the quality, specificity, and professional judgment that made their outputs valuable.
A cybersecurity consulting team based in the United Kingdom built their client engagements around diagnostic scenarios — structured simulations designed to assess organizational readiness for real-world threats. These scenarios were central to their service offering, but the process of creating them was entirely manual and time-intensive.
A single scenario could take up to two weeks to produce. Consultants would begin with drafting, then move through multiple rounds of client review, incorporate feedback, refine the content, and repeat the cycle until the scenario met the required standard. The iterative review process consumed a significant share of each engagement’s timeline — time that could otherwise be spent on actual analysis and advisory work.
As demand for their services grew, the manual content creation bottleneck became a direct constraint on the team’s capacity to take on new clients and deliver engagements efficiently. The firm needed a way to dramatically accelerate scenario production without sacrificing the quality, specificity, and professional judgment that made their outputs valuable.
Solution
Onix designed and built an AI-assisted workflow integrated into the consulting team’s existing platform, enabling consultants to generate high-quality cybersecurity scenarios in a fraction of the previous time — while keeping human expertise and final approval firmly in the process.
The system begins with a conversational AI agent that asks targeted clarifying questions to gather the specific organizational context needed before generating anything. Consultants can supplement this with uploaded reference documents and manual input to ensure outputs are grounded in the client’s actual environment.
Once sufficient context is collected, the system summarizes its understanding and presents several scenario options for the consultant to choose from. After selection, it generates a fully detailed scenario — including structured content and text-based “injects” — step by step, stored as drafts ready for consultant review and editing.
Critically, refinement now happens live, during the session itself, rather than through offline feedback rounds. This eliminates the multi-week review cycle and compresses iteration to a single working session. Data isolation is enforced per organization, and uploaded files are used only within the active session — never retained as a shared knowledge base.
The solution was built on Amazon Bedrock for AI services, with AWS infrastructure, S3 storage, and integrations with the client’s existing LMS and SSO systems.
Onix designed and built an AI-assisted workflow integrated into the consulting team’s existing platform, enabling consultants to generate high-quality cybersecurity scenarios in a fraction of the previous time — while keeping human expertise and final approval firmly in the process.
The system begins with a conversational AI agent that asks targeted clarifying questions to gather the specific organizational context needed before generating anything. Consultants can supplement this with uploaded reference documents and manual input to ensure outputs are grounded in the client’s actual environment.
Once sufficient context is collected, the system summarizes its understanding and presents several scenario options for the consultant to choose from. After selection, it generates a fully detailed scenario — including structured content and text-based “injects” — step by step, stored as drafts ready for consultant review and editing.
Critically, refinement now happens live, during the session itself, rather than through offline feedback rounds. This eliminates the multi-week review cycle and compresses iteration to a single working session. Data isolation is enforced per organization, and uploaded files are used only within the active session — never retained as a shared knowledge base.
The solution was built on Amazon Bedrock for AI services, with AWS infrastructure, S3 storage, and integrations with the client’s existing LMS and SSO systems.
Results
The AI-assisted workflow delivered a step-change improvement in the consulting team’s operational throughput:
70–80% reduction in scenario preparation time — from approximately two weeks down to a few days per scenario
Feedback cycles reduced from 2+ iterative rounds to a single live session in the majority of engagements
Significant reduction in manual consultant effort on content creation, freeing capacity for higher-value advisory work
Multiple scenario options can be generated and compared within a single working session
Faster client-facing turnaround, improving the overall responsiveness and competitiveness of the consulting team’s delivery model
The AI-assisted workflow delivered a step-change improvement in the consulting team’s operational throughput:
70–80% reduction in scenario preparation time — from approximately two weeks down to a few days per scenario
Feedback cycles reduced from 2+ iterative rounds to a single live session in the majority of engagements
Significant reduction in manual consultant effort on content creation, freeing capacity for higher-value advisory work
Multiple scenario options can be generated and compared within a single working session
Faster client-facing turnaround, improving the overall responsiveness and competitiveness of the consulting team’s delivery model