As we have been exploring, one of the powers of GenAI is the ability to automatically generate responses without being explicitly trained on it. Rather than just executing predefined tasks, LLMs can infer responses by drawing on their contextual understanding and knowledge. This aspect of emergent reasoning unlocks unique opportunities for rapid experimentation and iterative refinement of novel use cases.

When considering potential applications for GenAI, the first evaluation criterion centers on comprehension-based tasks. Sentiment analysis, content classification, intent classification, relationship extraction, summarization, and more all leverage innate language understanding. Developers can formulate prompts aligning to use cases that interpret, organize, or infer meaning. To unlock the full potential of LLMs, developers will iterate on these given prompts through thoughtful “prompt engineering.” Prompt engineering attempts to optimize...

When assessing opportunities to solve business problems with GenAI, it’s critical that we analyze potential value in business terms – how does this technology shift affect structures, efficiency, competitive positioning, or revenue opportunities in an organization? Merely showcasing sophisticated technical capabilities alone is not enough to show value, as it lacks strategic rigor. Proposed projects should directly address tangible problems or sources of organizational leverage.

As thought leaders seeking executive buy-in, we need to develop skills and methodologies that translate potential productivity gains into compelling business cases, clarifying the return on investment. Specifically, constructing cost avoidance models resonates with executives focused on operational efficiencies and margins.

Quantifying assumptions in terms that the business can understand and measure is an essential step to a successful integration of GenAI into...

As we evaluate potential applications for GenAI, two overarching categories emerge – comprehension applications and generative applications. By distinguishing use cases along this spectrum of understanding existing data vs synthesizing new artifacts, we can better match capabilities to appropriate business challenges:

• Comprehension applications cover processing tasks applied to analyze and structure knowledge from existing content. This includes sentiment analysis, relationship extraction, intent classification, summarization, and more. The key focus areas center around interpreting, organizing, and tagging data to be used by subsequent systems.

Sources like chat history, customer tickets, and product catalogs yield richer insights when machine learning models classify topics, normalize entities, and summarize concepts at scale.

• Generative applications move beyond pure comprehension applications to creatively...

When exploring opportunities to apply GenAI, continuously evaluate any potential business value first rather than just the technical art of the possible. To spark ideas, the following list summarizes promising use cases organized by key value drivers. Consider cases where replicating human-quality comprehension or creation at a machine scale has a positive impact on workflows.

Cost and efficiency gains:

• Automate high-volume/repetitive tasks. Here, your business key performance indicator (KPI) can be an automation rate, measured as a percentage of automated tasks.
• Accelerate content development, enabling hybrid human/AI collaboration. Here, your business KPI can be an increase in the number of hours saved compared to the current amount of time taken to create content.
• Content evaluation to reduce errors and rework. Here, your business KPI can be the delta in the number of tickets filed to update incorrect content.
...

As we consider which types of use cases we are looking to pursue to provide business value, we must consider the infrastructure on which we will deploy and host our systems. With the new normal of leveraging cloud resources, we tend to assume that capacity is not a concern anymore, but is this right? Let’s dissect this thought – is the biggest model the right solution for all use cases? Realistically speaking, LLMs are nice and easy to test and get initial results, but when considering scale and productionalization, they are not as appealing as you would think. Some of the limitations are GPU availability, cost, and latency. This realization is steering the market into more specialized smaller models that solve a specific use case.

Designing product architecture for LLMs requires careful consideration of several factors. Cost optimization strategies like Mixture-of-Depths can be employed to dynamically allocate resources...

This chapter explored the intricacies of evaluating potential use cases for GenAI, equipping you with the tools to discern optimal applications for this transformative technology. By examining different viewpoints and criteria, we’ve established a framework to determine whether a given use case aligns with GenAI’s strengths and limitations.

The key takeaways are as follows:

• At the heart of every successful GenAI implementation lies clear and demonstrable business value. Ask yourself how GenAI will directly enhance your current operations, unlock new revenue streams, or improve customer experience. Without a tangible benefit, the technology itself holds little merit.
• We identified two broad categories of use cases where GenAI shines: comprehension and generation. Comprehension focuses on analyzing and extracting meaning from data, while generation leverages that understanding to produce entirely new content. Consider which category your use...