关键字: [Amazon Web Services re:Invent 2023, Max.ai, Generative Ai Agents, Customer Experience, Ai Agents For Enterprises, Ai Agent Use Cases, Ai Agent Architecture]
Number of words: 1500, reading time: 8 minutes
video
If the video cannot be played normally, please go to bilibili to watch this video. >> https://www.bilibili.com/video/BV1bN411j7gD
Introduction
Generative AI assistants are transforming marketing content creation and customer contact experiences, while providing faster and better insights into the voice of the customer and responding to customer feedback. In this lightning talk, learn how enterprises can effectively leverage the power of Amazon Bedrock and Max.AI (a generative AI assistant platform) to deploy various AI assistants on a large scale, improve cross-functional productivity, and transform customer experience. This speech was provided by Amazon Cloud Technology Partner ZS.
Highlights of speech
The following is the essence of this speech compiled by the editor for you. It has a total of 1,200 words and takes about 6 minutes to read. If you want to know more about the content of the speech or watch the full text of the speech, please watch the full video of the speech or the original text of the speech below.
Over the past few months, speakers have noticed an increasing number of generative AI agents being deployed into production at multiple enterprises. When researching the latest trends, it is clear to see that generative AI is already being used in multiple areas such as marketing copywriting, brand creative, advertising, analytics, and video production. Especially the emerging GPT proxy, its powerful performance has been proven.
Now, many companies have been able to realize value quickly, taking only 2-3 months. The speaker emphasized that you can deploy generative AI directly in production and start realizing business value in less than three months. This is achieved through the max.ai platform they developed, which works with Amazon Cloud Technology’s Bedrock. This enables enterprises to create and deploy compliant AI agents within existing environments.
When looking at some of the use cases that emerged, in addition to basic Q&A functionality, it is also possible to provide valuable insights into ongoing issues impacting customer experience by unlocking the voice of the customer from different channels and analyzing sentiment. Speakers pointed out that max.ai can effectively provide enterprises with voice of customer analysis.
Content marketing is another impactful use case they see. Speakers will share a case study showing how to increase click-through rates and drive conversions for email and mobile campaigns.
These agents have also proven to be very powerful in training large numbers of employees and service personnel. For businesses with many reps, agents can determine the best training options and topics that resonate most, thereby increasing NPS scores in an efficient manner.
Additionally, customer service has always been a focus. When lower-level representatives encounter a new problem, they often spend a lot of time looking for a solution in documentation and other resources. GPT-driven agents can quickly surface the most relevant top solutions, increasing agent productivity.
While these are some prominent examples, speakers highlighted that many other deployable use cases in digital marketing and production can also generate solid returns.
Max.ai's success in enabling these use cases is largely due to its core agent architecture built on Amazon Cloud Technology. The company leverages advanced models such as GPT-3 and Anthropic, which are enterprise-grade certified through Amazon Cloud Technology’s infrastructure to meet regulatory compliance requirements. Whether an open source model or something else is needed for a specific use case, max.ai can make the best choice.
Once the model framework is ready, a key feature of the max.ai platform is the ability to ingest data from a variety of data sources within the enterprise and quickly organize them into a dynamic knowledge grid. This data will be converted into vector embeddings to drive the agent. In the past, the process of ingesting disparate enterprise data sources has been painful, but max.ai has innovated to significantly speed up the process, the spokesperson said.
Next, the LM inference process occurs. Additionally, other key features such as long-term memory, additional tools to prevent phantom content protection, and integration connectors with required enterprise applications are provided.
When discussing generative AI agents, speakers outlined their vision for the prospects for such applications. At the core is the LM that drives the agents, but more importantly leverages traditional AI and automation technologies to connect the agents with the enterprise’s existing systems. This enables agents to perform digital tasks within these systems, such as sending campaigns in Salesforce, executing marketing campaigns in Adobe, or handling customer service tasks in Zendesk.
The key is, all of this is provided to business users in an easy-to-use interface, or via API endpoints. So for organizations that are running legacy applications, they can plug these agents into those systems to immediately take advantage of the benefits of large language models.
Overall, this is an approach that allows for the deployment of generative AI agents to Amazon Cloud infrastructure in approximately 6-8 weeks. The spokesperson emphasized that these proxies are very easy to set up, making it possible to leverage the Amazon Cloud Marketplace and connect them to other systems already in place within the enterprise. Therefore, it is entirely feasible from a technical and business perspective to push these agents into production.
In a client-specific case study, the speaker detailed one agency's process for conducting email subject line testing and marketing copy testing at scale. This client sends nearly 8 million emails per week, and the agency has achieved double-digit growth through improved click-through and conversion rates. Forrester has recently confirmed these results.
The agency uses a variety of marketing tools—event details, traditional marketing content, voice patterns, sentiment scores, and more. With the power of large language models, agencies can generate a wide variety of subject line variations for testing. However, this is just the beginning. Next, the agency identifies customer groups to test, combines LM-generated content with the experimentation framework, and pushes integration campaigns through platforms such as Adobe Target.
The differentiating power that agencies demonstrate lies in the psychographic insights they reveal about the ideal tone of voice, subject line style, and content preferences of their customer base. This insight brings unprecedented behavioral data to enterprise systems to further optimize content.
The speaker listed a range of pre-built agency options that companies can deploy. For example, Maverick handles Q&A across documents and multi-modal use; Raven focuses on voice of customer analysis; Kit covers customer complaints, customer success, and agent productivity; and content marketing agencies specialize in generative AI technology.
For example, Maverick can improve the productivity of knowledge tasks by 20-25% by querying ingested enterprise data. It provides user access control to collection and connection sources such as S3. Kit performs sentiment analysis on customer complaints to uncover key insights. Instead of regular NLP projects, users can customize the topics they want to monitor, so Kit can continually report new issues.
Content marketing agencies incorporate generative AI to create content variations and test experimental frameworks on customer groups. This helps enable data-driven content optimization.
The speakers emphasized that with the rapid development of large language models, they are working towards a feature called "build your own agents." This feature enables enterprise users to easily create custom agents and run them under governance and security protocols. This toolkit uses selected models such as GPT-4, provides configurable functionality from QA to chatbots, and can be integrated into any department.
Customers have created over 100 unique agents for business users within their organization. The goal is to commoditize basic use cases to enable customizable self-service agents. This kind of DIY agent is affordable and not limited by the number of agents. It can be used in many aspects such as marketing, supply chain, safety, and product documentation. Advanced use cases require deeper customization.
Overall, the key takeaway is that generative AI agents can create value for businesses in different areas such as marketing, customer service, training and content creation, in just 2-3 months. The max.ai platform is based on Amazon Cloud Technology's infrastructure and provides configurable deployment of compliant AI agents. Pre-built proxies are available for common use cases, or customizable DIY proxies can be created to commoditize basic functionality. This will democratize access to generative AI, bringing these powerful capabilities to the doorstep of enterprises in a way that is easy for them to use.
Here are some highlights from the speech:
Leaders discussed generative agent technology used by multiple companies in production.
With the help of Amazon Cloud Technology Platform, enterprises can quickly deploy AI solutions and realize business value in just three months.
Amazon Cloud Technologies provides customers with demographic insights to help them craft more targeted marketing content.
These generative AI agents can analyze customer feedback to identify key concerns, such as product features and security issues.
In addition, the leader has launched a new Enterprise Edition product on the Amazon Cloud Technology Marketplace for businesses to download and use in customers' own environments.
Summarize
This speech mainly discusses how enterprises can use artificial intelligence agents to improve customer experience. The speakers highlighted that their company has developed a range of generative AI agents suitable for a variety of scenarios such as marketing, customer service and employee training. According to the speaker, these AI agents can be put into production within 2 to 3 months, thereby creating value. These agents are built on Amazon cloud technology and employ models such as GPT-4 and Anthropic to ensure enterprise governance and compliance. Some key application scenarios include:
- User voice analysis to understand customer emotions and needs. The AI agent continuously monitors various data sources to identify new issues and problems.
- Content marketing to generate high-quality subject lines and copy. AI agents are able to test content at scale and adapt it to specific customer groups. This initiative has resulted in double-digit increases in email click-through rates and customer conversion rates.
- Customer service automation to quickly analyze problems and find solutions from the knowledge base. This improves agent productivity.
The company offers pre-built agents such as Maverick for QA and Raven for user speech analysis. They also allow businesses to easily customize AI agents for various departments through a no-code interface. Their goal is for companies to have an "AI army" across the entire enterprise.
In summary, this presentation demonstrated the rapid development of generative AI and how enterprises can improve customer and employee experiences through customizable AI agents.
Original speech
https://blog.csdn.net/just2gooo/article/details/134868273
Want to know more exciting and complete content? Visit re:Invent official Chinese website now!
2023 Amazon Cloud Technology re:Invent Global Conference - Official Website
Click here to get the latest global product/service information from Amazon Cloud Technology with one click!
Click here to get the latest product/service information from Amazon Cloud Technology China with one click!
Register an Amazon Cloud Technology account now and start your cloud journey!
[Free] Amazon Cloud Technology "Free trial of more than 100 core cloud service products"
[Free] "Free trial of more than 40 core cloud service products" of Amazon Cloud Technology China
Who is Amazon Cloud Technology?
Amazon Cloud Technology (Amazon Web Services) is the pioneer and leader of global cloud computing. Since 2006, it has been characterized by continuous innovation, technology leadership, rich services, and wide application And well-known in the industry. Amazon Cloud Technology can support almost any workload on the cloud. Amazon Cloud Technology currently provides more than 200 full-featured services, covering computing, storage, network, database, data analysis, robotics, machine learning and artificial intelligence, Internet of Things, mobile, security, hybrid cloud, virtual reality and augmented reality, media , as well as application development, deployment and management; the infrastructure covers 99 availability zones in 31 geographical regions, and plans to build 4 new regions and 12 availability zones. Millions of customers around the world, from startups, small and medium-sized enterprises, to large enterprises and government agencies, trust Amazon Cloud Technology. They use Amazon Cloud Technology services to strengthen their infrastructure, improve agility, reduce costs, accelerate innovation, and enhance competitiveness. Achieve business growth and success.
