{"id":24237,"date":"2024-03-25T07:46:03","date_gmt":"2024-03-25T14:46:03","guid":{"rendered":"https:\/\/3cloudsolutions.com\/?p=24237"},"modified":"2024-07-09T08:14:04","modified_gmt":"2024-07-09T15:14:04","slug":"generative-ai-assessment-client-example","status":"publish","type":"post","link":"https:\/\/3cloudsolutions.com\/resources\/generative-ai-assessment-client-example\/","title":{"rendered":"Generative AI Assessment: Client Example"},"content":{"rendered":"

Artificial Intelligence<\/a> (AI) is hot topic and organizations across every industry are looking for ways to successfully integrate it into their business. The benefits of AI for businesses include increased efficiency through automation, enhanced decision making with data, improved customer experiences through personalized interactions, and the ability to uncover valuable insights from large datasets that would be impossible for humans to manually analyze. Additionally, AI can enable businesses to identify market trends, optimize processes, and continue fostering innovation. Companies understand that they need to embrace AI or be left behind, although most don\u2019t know how to begin.<\/span><\/p>\n

3Cloud\u2019s generative AI<\/a> assessment helps businesses to identify their starting point with AI through a phased progression of education, ideation, and proof of value to achieve scale across the business. This blog walks through a construction architecture and engineering firm\u2019s AI assessment engagement with 3Cloud.<\/span><\/p>\n

AI Roadmap Engagement<\/span><\/h3>\n

The client partnered with 3Cloud to collaborate and perform an AI Roadmap engagement, aiming to leverage the potential of artificial intelligence within their organization. Throughout the assessment, several key objectives were identified and prioritized. These included determining whether to build or buy generative AI solutions<\/a>, outlining a high-level AI architecture framework with considerations for maintenance, devising strategies for efficient retrieval and prompt-engineering, improving the accuracy of Q&A systems over arbitrarily large knowledge bases, and evaluating and assessing various Retrieval-Augmented Generation (RAG) architectures.<\/span><\/p>\n

Existing Capabilities and Restraints<\/span><\/h3>\n

In assessing the existing capabilities of the organization, their constraints became apparent, particularly regarding the limited number of Enterprise Technology Subject Matter Experts (SMEs), which restricts comprehensive support across various technical domains. Also, the cost-prohibitive nature of fully staffing the Service Desk poses a challenge, necessitating a more efficient allocation of resources. A third constraint was the difficulty in managing knowledge for large-scale construction projects.<\/span><\/p>\n

Transitioning to the desired state involves implementing GenAI solutions such as a “Chat with an IT Architect\u201d feature to streamline access to expert guidance, augmenting Service Desk staff with AI to handle routine inquiries and refocus roles on advanced triage, and establishing a solution to make the project knowledge accessible to all participants within seconds. These initiatives aim to enhance operational efficiency, optimize resource utilization, and improve collaboration across the organization.<\/span><\/p>\n

Use Case: Chat with an IT Architect<\/span><\/p>\n

The primary use case identified was to create an AI solution to \u201cChat with an IT Architect\u201d. The goal was to ensure a cohesive data strategy for Retrieval-Augmented Generation (RAG) across multiple knowledge bases (KBs). Architecting and budgeting for this use case would involve allocating resources to key components such as data storage and transformations, a modern data platform<\/a>, reporting and BI tools<\/a>, and document cracking technologies.<\/span><\/p>\n

When considering the architecture and management of data for this use case, several factors come into play. Firstly, data governance practices<\/a> are essential to ensure organized indexing of data types for efficient retrieval, segmentation and categorization. Next, maintaining high data quality is crucial, following the principle of GIGO (garbage-in, garbage-out). Thirdly, implementing data version control tools enables tracking changes and ensuring data freshness, particularly vital for knowledge base updates. Lastly, unified orchestration of data workflows and pipelines ensures\u00a0 efficiency and effectiveness in leveraging data for AI-driven insights.<\/span><\/p>\n

Use Case: Augment Service Desk<\/span><\/h3>\n

The other use case identified was to augment Service Desk staff with AI that can engage customers. When managing data for AI applications, several key considerations should be kept in mind. In this case, implementing data version control to track changes and iterations in data sources, ensuring accurate QA to maintain consistency, i.e., similar ticket resolutions or the latest updates. Next, establishing clear governance policies is crucial for ensuring data integrity, privacy, and compliance. Last, addressing multi-modality challenges by handling various data types such as text, images, etc., that may complicate data pre-processing but can greatly improve groundedness.\u00a0<\/span><\/p>\n

How will the Listing Augmentation use-case impact the bottom line?<\/span><\/h4>\n

By advancing the construction company\u2019s AI journey and executing the identified use cases, the business will see significant impacts:<\/span><\/p>\n