Realize business value by transforming data into action with generative AI

May 3, 2024 | 5 minute read
Barry Mostert
Senior Director, Artificial Intelligence and Analytics
Text Size 100%:

In today’s data-driven world, businesses are constantly seeking innovative ways to harness the power of AI to extract insights, streamline processes, and drive decision-making. AI offers a wide range of capabilities across various stages of the data lifecycle, from generation and extraction to transformation and reasoning. Understanding these capabilities and how they can be applied in business contexts is essential for unlocking AI’s full potential.

Chart of prebuilt and fine-tuned large language models.
Figure 1: The four general categories of generative AI use cases.

Generation and creativity: Creating new possibilities

In the realm of generation and creativity, AI excels at producing novel content and creative outputs based on prompts or data input, including code generation capabilities like Oracle’s Select AI that automatically generates SQL syntax from natural language queries. Natural language generation (NLG) is another application that generates human-like text or speech for use cases like automated report creation, personalized content like dynamically crafting targeted sales emails with relevant product offerings for each customer and enabling conversational AI interfaces to otherwise complex corporate systems. By tapping into AI’s generative abilities, businesses can streamline development, enhance personalization, and boost productivity.  The following gif shows the process of generative AI combined with retrieval-augmented generation (RAG) that optimizes the corporate search experience. Instead of employees sifting through files hoping to find buried answers, this AI capability allows natural language queries to instantly retrieve and generate relevant answers.

Optimizing search with generative AI and retrieval-augmented generation.
Figure 2: Optimizing search with generative AI and retrieval-augmented generation.


Oracle employees struggle to find answers within vast amounts of internal microsites and documents, such as IT and HR. When answers aren’t found IT tickets (SRs) or emails to HR are generated asking for help placing load on HR and IT.

As a solution, MyOracle Search uses OCI Generative AI plus RAG to enable natural language interactions to find answers fast. Answers are provided directly inline and are grounded with references to original source materials. As a result, employees are more productive and self-sufficient, and 25–30% of common IT requests are answered by generative AI allowing IT agents to focus on complex tickets.

Summarization and extraction: Distilling insights from data

Summarization and extraction involves distilling large amounts of information into shorter, more concise forms while preserving key points or essential information. For example, topic detection uses AI to automatically identify and extract underlying themes or topics present in a collection of documents or text data, enabling tasks such as document classification and trend analysis. Extractive summarization is another technique where AI identifies and extracts the most important or relevant sentences or passages from a document to create a concise summary, making it useful for quickly understanding the main points of lengthy documents or articles.  For complex documents laden with industry jargon, like insurance policies in the following example, generative AI can extract and summarize the key points in a more comprehensible way for the reader.

An example insurance policy with information extracted by generative AI.
Figure 3: An example insurance policy with information extracted by generative AI.


Finding the answers within mountains of documents takes time and critical information might be inadvertently missed. But GenAI can be trained on large amounts of information in a variety of formats. Generate accurate summaries in supported languages to any level of brevity. As a result, you reduce the human time to digest information.

Rephrasing and transforming: Adapting to different needs

Rephrasing and transforming encompasses tasks, such as rewriting existing content while retaining the original meaning or reformatting documents into different formats. For example, translation involves using AI to convert text from one language to another while accurately conveying the meaning and nuances of the original content. This capability plays a crucial role in facilitating communication and information exchange in multicultural and multilingual environments, including international business, diplomacy, and content localization. In the following example, a customer service response is written with unintended negative sentiment that could damage the experience. Generative AI can intercept the response and rephrase it with a more positive, solution-oriented tone.

An example of rephrasing a customer service response
Figure 4: An example of rephrasing a customer service response.


Humans can inadvertently introduce ambiguity, unintended tone, or mistakes into their communications. But they also write their communications focusing on facts, not style or tone. Generative AI transforms the human input, correcting for mistakes, corporate style. and technical information. Better communications improve customer experience. As a result, authors gain efficiency by reducing time wordsmithing communications.

Reason and act: Turning insights into action

The 'reason and act' category showcases AI’s proficiency at inference tasks like answering queries and making predictions from data. It doesn’t imply human-like reasoning, but rather systematic problem detection, diagnosis, and resolution. Examples include multi-turn conversational interfaces to solve complex issues like supply chain disruptions, shown below, through logical inferences, and root cause analysis dialogues where AI recommends resolution steps based on the inputs. While AI can bring answers, best practice today is for humans to validate outputs before final decisions, limiting full automation but building trust in AI’s recommendations over time.

An exmample of an SCM assistant help diagnose and resolve a problem.
Figure 5: An exmample of an SCM assistant help diagnose and resolve a problem.


Because disruptions are costly, we must monitor the supply chain in real time, including external data, and gain a conversational interface to logically break down the problem and evaluate potential solutions. It’s like chatting with a helpful and informed colleague. The process gives rapid response to potential disruptions so that we can evaluate more solutions.

Embracing AI for business transformation

As businesses continue to navigate the complexities of the digital age, AI emerges as a powerful ally in driving innovation, efficiency, and competitiveness. By using AI’s capabilities across the data lifecycle—from generation and extraction to transformation and reasoning—organizations can unlock new opportunities, streamline operations, and make informed decisions. Embracing AI for business transformation isn’t just about adopting the latest technologies. It's about embracing a mindset of continuous learning, experimentation, and adaptation in the pursuit of excellence.

Are you ready to embark on your generative AI journey? Try the following resources to help you choose your first use case:

Barry Mostert

Senior Director, Artificial Intelligence and Analytics

Barry is a senior director for product marketing covering Oracle's AI and Analytics services.

Previous Post

Introducing Meta Llama 3 on OCI Ampere A1: A testament to CPU-Based Model Inference

Amar Gowda | 4 min read

Next Post

Using ML Insights for observability at cloud scale

Antariksha Bhaduri | 10 min read