ASRock Industrial Transforms Technical Support with AI TSD Agent on AI BOX-A395 and AiKMSpark
Technical support plays a critical role in helping customers resolve product issues while preserving the engineering knowledge gained along the way. At ASRock Industrial, the Technical Support Division (TSD) manages a large and growing collection of support cases, engineering reports, troubleshooting records, and product knowledge built up over years of customer interactions. This accumulated knowledge is a valuable resource for problem analysis, root cause investigation, and ongoing support. To make better use of these assets and put Enterprise AI to practical use, ASRock Industrial deployed AI BOX-A395 with AiKMSpark as an on-premises enterprise AI knowledge management platform. Built on this core solution, ASRock Industrial further developed the AI TSD Agent around the specific workflows and knowledge requirements of its Technical Support Division (TSD). The deployment demonstrates how AI BOX-A395 and AiKMSpark can serve as a foundation for tailored enterprise AI applications, turning accumulated organizational knowledge into practical AI-powered tools while keeping proprietary data entirely within the company’s infrastructure.
Challenge
Every support case adds to ASRock Industrial’s body of engineering knowledge. Over the years, the Technical Support Division (TSD) has accumulated extensive technical cases and support records covering troubleshooting, root cause analysis, and proven resolutions across thousands of customer interactions. Before adopting an AI-powered knowledge system, engineers primarily relied on a technical case database and historical emails to search for past solutions.
As the volume of cases grew, TSD faced three key challenges:
- Limited Keyword-Based Search：Relevant cases could be difficult to find when search terms were not precise or did not closely match the original case content.
- Time-Consuming Manual Review：Engineers had to manually review case details and historical emails, with limited summaries or connections between related issues.
- Difficult Knowledge Discovery for New Engineers：New engineers often found it challenging to identify the right information and learn from relevant past cases.
As customer issues became increasingly complex, ASRock Industrial recognized the need for a more intelligent way to search, summarize, connect, and reuse engineering knowledge, while keeping proprietary technical data securely under the company’s control.
Solution
AI BOX-A395 + AiKMSpark: The Foundation for Enterprise AI
To address this challenge, ASRock Industrial deployed AI BOX-A395 with AiKMSpark to establish a secure, on-premises AI knowledge platform for its Technical Support Division (TSD). AI BOX-A395 provides the local AI computing foundation, while AiKMSpark transforms years of accumulated technical cases into a structured and searchable knowledge base. AiKMSpark reads imported enterprise documents and organizes their content into a structured knowledge base, without requiring engineers to define every knowledge category in advance. Engineers can access this accumulated knowledge through natural language, describing an issue directly instead of relying on keyword search. Because AiKMSpark runs locally on AI BOX-A395, the source documents, knowledge base, and AI models stay within ASRock Industrial’s own infrastructure, giving the company full control over its proprietary engineering data.
Tailored Development: From AiKMSpark to AI TSD Agent
With this foundation in place, ASRock Industrial tailored AiKMSpark to the specific knowledge and workflow of its TSD, creating the AI TSD Agent. Historical technical cases – spanning engineering reports, text files, spreadsheets, and other support records accumulated over years of customer interactions, were imported into the system. AI TSD Agent reads through this material and extracts the details that matter for troubleshooting, including products, issues, root causes, and resolutions, organizing years of accumulated technical experience into a knowledge base engineers can actually use. When a new issue comes in, engineers describe it in natural language instead of paging back through past cases individually. AI TSD Agent retrieves the historical cases that match, summarizes the key findings, identifies similar issues, suggests possible troubleshooting approaches, and provides supporting references for the engineer’s own analysis. It also helps with root cause report preparation, organizing relevant findings into a draft report that engineers review and finalize. As new cases are closed, they are added back into the knowledge base, so the accumulated experience keeps growing and stays available for the next case that comes in.
Why AI BOX-A395 + AiKMSpark
The AI BOX-A395 provides the local computing foundation required to bring enterprise knowledge management and AI workloads directly into the organization. Powered by the AMD Ryzen™ AI Max+ 395 processor and supporting up to 128GB of unified memory, the AI BOX-A395 delivers up to 126 TOPS of overall AI performance, providing the computing resources required to run large language models locally while processing extensive enterprise knowledge. Combined with AiKMSpark, organizations can build and maintain enterprise knowledge bases without relying on cloud-based AI services. Documents and AI models can remain within the organization’s own infrastructure, while local deployment provides greater control over data, AI workloads, and operating costs. The deployment of AI TSD Agent within ASRock Industrial demonstrates how this foundation can be tailored to specific enterprise workflows and knowledge requirements, turning enterprise knowledge into a practical AI application.
Benefits
Workflow Optimization:
AI TSD Agent streamlines issue analysis, root cause reporting, and knowledge accumulation, optimizing approximately 50% (three of the six stages) of the overall technical support workflow.
Faster Knowledge Retrieval:
Quickly surfaces relevant historical cases, summarizes key findings, and suggests possible solutions, reducing the time required for technical analysis and customer support.
Improved Technical Support Efficiency:
Natural language search and AI-assisted report generation reduce repetitive tasks, enabling engineers to analyze and respond to technical issues 10–20% more efficiently.
Continuous Knowledge Growth:
Organizes accumulated technical knowledge into a structured knowledge base that expands with every completed support case, making engineering experience easier to preserve and reuse.
Secure On-Premises Deployment:
Enterprise documents, customer information, and AI models remain within the organization’s own infrastructure instead of being sent to external cloud AI services.
Adaptable Enterprise AI Foundation:
The combination of AI BOX-A395 and AiKMSpark provides a foundation that can be tailored to different enterprise knowledge requirements, with AI TSD Agent demonstrating how the platform can be developed into a practical AI application for a specific business workflow.