Article by Priya Arora, Global Head, AWS Generative AI Center of Excellence
by Jacob Newton-Gladstein, Field Enablement Lead, Generative AI Center of Excellence – AWS
As the field of generative AI continues to evolve, companies are looking for ways to evaluate generative AI prototypes and determine which ones create real business value, which means establishing a tangible link between an application and its overall business value.
The latest assets available through the Generative AI Center of Excellence (CoE) are focused on helping AWS Partners understand their business opportunity by engaging with Generative AI thought leaders and CoE contributors to deep dive into use cases. Visit the CoE to learn more about LLM Precision in Biomedical Engineering from Provectus and Generative AI for Retail and Ecommerce from Software One.
Additionally, the AWS Generative AI Center of Excellence worked with QuantumBlack and AI by McKinsey to provide AWS Partners with insights into the unique needs and priorities of different personas working on the topic of generative AI.
Because business value is derived differently across different pillars of an organization, we explore potential use cases for generative AI across four different organizational business units, which represent 75% of the total opportunity for generative AI workloads according to McKinsey research. Specifically, we dive deep into the needs of operations leaders, research and development (R&D) leaders, revenue leaders, and marketing leaders.
Key Considerations When Engaging Revenue Leaders
Revenue leaders are uniquely positioned for generative AI because they span both customer operations and marketing and sales functions: McKinsey & Company estimates that generative AI could improve sales productivity by roughly 3 to 5 percent of current global sales spend.
Revenue growth leaders want a very clear correlation between generative AI and revenue growth, so consider setting specific growth goals for your generative AI workloads to achieve. Finally, focusing on reusable base application components that can be adapted across use cases can shorten delivery timelines.
Key considerations when engaging with marketing leaders
Marketing leaders have a huge opportunity to improve their organizations through generative AI: McKinsey & Company estimates that generative AI can increase productivity in the marketing function, with the value reaching 5 to 15 percent of total marketing spend.
To take advantage of these opportunities, consider the metrics CMOs care about: conversion rates, customer retention, and ad spend. To effectively demonstrate the value of generative AI to marketing leaders, consider building tools and dashboards to A/B test personalized ads and ad spend, and measure conversion rates in tandem with your generative AI solution.
Generative AI for R&D Leaders
R&D leaders can leverage generative AI to advance their entire value chain: McKinsey & Company estimates that generative AI can deliver productivity gains worth 10 to 15 percent of overall R&D costs.
R&D leaders care not only about business outcomes, but also about data accuracy, explainability, citation, data privacy, etc. To appeal to these leaders, focus on building tools that establish clear links between data inputs and outputs, define clear pre- and post-processing rules, and improve transparency of the outcomes of your generative AI tools.
Key Considerations When Engaging Operations Leaders
Because increased productivity is a major benefit of generative AI, operations leaders are well positioned to adopt it: McKinsey & Company estimates that generative AI could add up to $280 billion to $530 billion in value to operational functions across industries.
When designing generative AI applications for operational leaders, keep in mind that large-scale operations incorporate vast amounts of data and a diverse operator base. As a result, a well-thought-out data map and proactive integration plan for how generative AI applications will connect to the organization's existing work processes will be helpful.
How Generative AI on AWS enables partners to reach cross-industry leaders
All generative AI decision makers have some common needs.
- Generative AI applications go beyond large-scale language models — you need to understand the entire generative AI stack.
- Different problems require different levels of solutions, ranging from purpose-built applications to custom-trained models and even adapted models through Amazon Bedrock.
- It is easier to prototype generative AI applications than to deploy them in production to deliver sustained business value.
By helping customers build generative AI solutions on AWS, organizations can access a wide range of generative AI tools, including multiple options for generative AI data structures such as data lakes and the Amazon OpenSearch Services Vector Database, to prepare data for generative AI applications.
You can also access the broadest list of adaptive LLMs through Amazon SageMaker JumpStart and reduce your model training costs by leveraging AWS's purpose-built generative AI chips, AWS Trainium and AWS Inferentia.
Amazon Bedrock makes it easy to move to production. Customers can quickly leverage data already available, choose the best FM, establish guardrails, and rest assured with the security and reliability of AWS. With Amazon Q, you have access to multiple products tailored to different personas and use cases. You can use Amazon Q to answer employee questions, summarize data, continue conversations, and drive action. It can also lighten the load on developers by helping them write, debug, test, and transform code.
Finally, the AWS Generative AI Center of Excellence also features thought leadership articles from AWS Partners, such as “LLM Accuracy in Biomedical Engineering” by Provectus and “Generative AI for Retail and Ecommerce” by SoftwareOne, that can help guide you through use-case specific applications of Generative AI across industries and enable you to move your applications from POC to production.
Leverage the Generative AI Center of Excellence to get resources to help you build and execute your Generative AI go-to-market strategy (login required).
