Post Syndicated from Antje Barth original https://aws.amazon.com/blogs/aws/build-generative-ai-applications-with-amazon-bedrock-studio-preview/

Today, we’re introducing Amazon Bedrock Studio, a new web-based generative artificial intelligence (generative AI) development experience, in public preview. Amazon Bedrock Studio accelerates the development of generative AI applications by providing a rapid prototyping environment with key Amazon Bedrock features, including Knowledge Bases, Agents, and Guardrails.

As a developer, you can now use your company’s single sign-on credentials to sign in to Bedrock Studio and start experimenting. You can build applications using a wide array of top performing models, evaluate, and share your generative AI apps within Bedrock Studio. The user interface guides you through various steps to help improve a model’s responses. You can experiment with model settings, and securely integrate your company data sources, tools, and APIs, and set guardrails. You can collaborate with team members to ideate, experiment, and refine your generative AI applications—all without requiring advanced machine learning (ML) expertise or AWS Management Console access.

As an Amazon Web Services (AWS) administrator, you can be confident that developers will only have access to the features provided by Bedrock Studio, and won’t have broader access to AWS infrastructure and services.

Now, let me show you how to get started with Amazon Bedrock Studio.

Get started with Amazon Bedrock Studio
As an AWS administrator, you first need to create an Amazon Bedrock Studio workspace, then select and add users you want to give access to the workspace. Once the workspace is created, you can share the workspace URL with the respective users. Users with access privileges can sign in to the workspace using single sign-on, create projects within their workspace, and start building generative AI applications.

Create Amazon Bedrock Studio workspace
Navigate to the Amazon Bedrock console and choose Bedrock Studio on the bottom left pane.

Before creating a workspace, you need to configure and secure the single sign-on integration with your identity provider (IdP) using the AWS IAM Identity Center. For detailed instructions on how to configure various IdPs, such as AWS Directory Service for Microsoft Active Directory, Microsoft Entra ID, or Okta, check out the AWS IAM Identity Center User Guide. For this demo, I configured user access with the default IAM Identity Center directory.

Next, choose Create workspace, enter your workspace details, and create any required AWS Identity and Access Management (IAM) roles.

If you want, you can also select default generative AI models and embedding models for the workspace. Once you’re done, choose Create.

Next, select the created workspace.

Then, choose User management and Add users or groups to select the users you want to give access to this workspace.

Back in the Overview tab, you can now copy the Bedrock Studio URL and share it with your users.

Build generative AI applications using Amazon Bedrock Studio
As a builder, you can now navigate to the provided Bedrock Studio URL and sign in with your single sign-on user credentials. Welcome to Amazon Bedrock Studio! Let me show you how to choose from industry leading FMs, bring your own data, use functions to make API calls, and safeguard your applications using guardrails.

Choose from multiple industry leading FMs
By choosing Explore, you can start selecting available FMs and explore the models using natural language prompts.

If you choose Build, you can start building generative AI applications in a playground mode, experiment with model configurations, iterate on system prompts to define the behavior of your application, and prototype new features.

Bring your own data
With Bedrock Studio, you can securely bring your own data to customize your application by providing a single file or by selecting a knowledge base created in Amazon Bedrock.

Use functions to make API calls and make model responses more relevant
A function call allows the FM to dynamically access and incorporate external data or capabilities when responding to a prompt. The model determines which function it needs to call based on an OpenAPI schema that you provide.

Functions enable a model to include information in its response that it doesn’t have direct access to or prior knowledge of. For example, a function could allow the model to retrieve and include the current weather conditions in its response, even though the model itself doesn’t have that information stored.

Safeguard your applications using Guardrails for Amazon Bedrock
You can create guardrails to promote safe interactions between users and your generative AI applications by implementing safeguards customized to your use cases and responsible AI policies.

When you create applications in Amazon Bedrock Studio, the corresponding managed resources such as knowledge bases, agents, and guardrails are automatically deployed in your AWS account. You can use the Amazon Bedrock API to access those resources in downstream applications.

Here’s a short demo video of Amazon Bedrock Studio created by my colleague Banjo Obayomi.

Join the preview
Amazon Bedrock Studio is available today in public preview in AWS Regions US East (N. Virginia) and US West (Oregon). To learn more, visit the Amazon Bedrock Studio page and User Guide.

Give Amazon Bedrock Studio a try today and let us know what you think! Send feedback to AWS re:Post for Amazon Bedrock or through your usual AWS contacts, and engage with the generative AI builder community at community.aws.

— Antje

Post Syndicated from Esra Kayabali original https://aws.amazon.com/blogs/aws/guardrails-for-amazon-bedrock-now-available-with-new-safety-filters-and-privacy-controls/

Today, I am happy to announce the general availability of Guardrails for Amazon Bedrock, first released in preview at re:Invent 2023. With Guardrails for Amazon Bedrock, you can implement safeguards in your generative artificial intelligence (generative AI) applications that are customized to your use cases and responsible AI policies. You can create multiple guardrails tailored to different use cases and apply them across multiple foundation models (FMs), improving end-user experiences and standardizing safety controls across generative AI applications. You can use Guardrails for Amazon Bedrock with all large language models (LLMs) in Amazon Bedrock, including fine-tuned models.

Guardrails for Bedrock offers industry-leading safety protection on top of the native capabilities of FMs, helping customers block as much as 85% more harmful content than protection natively provided by some foundation models on Amazon Bedrock today. Guardrails for Amazon Bedrock is the only responsible AI capability offered by top cloud providers that enables customers to build and customize safety and privacy protections for their generative AI applications in a single solution, and it works with all large language models (LLMs) in Amazon Bedrock, as well as fine-tuned models.

Aha! is a software company that helps more than 1 million people bring their product strategy to life. “Our customers depend on us every day to set goals, collect customer feedback, and create visual roadmaps,” said Dr. Chris Waters, co-founder and Chief Technology Officer at Aha!. “That is why we use Amazon Bedrock to power many of our generative AI capabilities. Amazon Bedrock provides responsible AI features, which enable us to have full control over our information through its data protection and privacy policies, and block harmful content through Guardrails for Bedrock. We just built on it to help product managers discover insights by analyzing feedback submitted by their customers. This is just the beginning. We will continue to build on advanced AWS technology to help product development teams everywhere prioritize what to build next with confidence.”

In the preview post, Antje showed you how to use guardrails to configure thresholds to filter content across harmful categories and define a set of topics that need to be avoided in the context of your application. The Content filters feature now has two additional safety categories: Misconduct for detecting criminal activities and Prompt Attack for detecting prompt injection and jailbreak attempts. We also added important new features, including sensitive information filters to detect and redact personally identifiable information (PII) and word filters to block inputs containing profane and custom words (for example, harmful words, competitor names, and products).

Guardrails for Amazon Bedrock sits in between the application and the model. Guardrails automatically evaluates everything going into the model from the application and coming out of the model to the application to detect and help prevent content that falls into restricted categories.

You can recap the steps in the preview release blog to learn how to configure Denied topics and Content filters. Let me show you how the new features work.

New features
To start using Guardrails for Amazon Bedrock, I go to the AWS Management Console for Amazon Bedrock, where I can create guardrails and configure the new capabilities. In the navigation pane in the Amazon Bedrock console, I choose Guardrails, and then I choose Create guardrail.

I enter the guardrail Name and Description. I choose Next to move to the Add sensitive information filters step.

I use Sensitive information filters to detect sensitive and private information in user inputs and FM outputs. Based on the use cases, I can select a set of entities to be either blocked in inputs (for example, a FAQ-based chatbot that doesn’t require user-specific information) or redacted in outputs (for example, conversation summarization based on chat transcripts). The sensitive information filter supports a set of predefined PII types. I can also define custom regex-based entities specific to my use case and needs.

I add two PII types (Name, Email) from the list and add a regular expression pattern using Booking ID as Name and [0-9a-fA-F]{8} as the Regex pattern.

I choose Next and enter custom messages that will be displayed if my guardrail blocks the input or the model response in the Define blocked messaging step. I review the configuration at the last step and choose Create guardrail.

I navigate to the Guardrails Overview page and choose the Anthropic Claude Instant 1.2 model using the Test section. I enter the following call center transcript in the Prompt field and choose Run.

Please summarize the below call center transcript. Put the name, email and the booking ID to the top:
Agent: Welcome to ABC company. How can I help you today?
Customer: I want to cancel my hotel booking.
Agent: Sure, I can help you with the cancellation. Can you please provide your booking ID?
Customer: Yes, my booking ID is 550e8408.
Agent: Thank you. Can I have your name and email for confirmation?
Customer: My name is Jane Doe and my email is [email protected]
Agent: Thank you for confirming. I will go ahead and cancel your reservation.

Guardrail action shows there are three instances where the guardrails came in to effect. I use View trace to check the details. I notice that the guardrail detected the Name, Email and Booking ID and masked them in the final response.

I use Word filters to block inputs containing profane and custom words (for example, competitor names or offensive words). I check the Filter profanity box. The profanity list of words is based on the global definition of profanity. Additionally, I can specify up to 10,000 phrases (with a maximum of three words per phrase) to be blocked by the guardrail. A blocked message will show if my input or model response contain these words or phrases.

Now, I choose Custom words and phrases under Word filters and choose Edit. I use Add words and phrases manually to add a custom word CompetitorY. Alternatively, I can use Upload from a local file or Upload from S3 object if I need to upload a list of phrases. I choose Save and exit to return to my guardrail page.

I enter a prompt containing information about a fictional company and its competitor and add the question What are the extra features offered by CompetitorY?. I choose Run.

I use View trace to check the details. I notice that the guardrail intervened according to the policies I configured.

Now available
Guardrails for Amazon Bedrock is now available in US East (N. Virginia) and US West (Oregon) Regions.

For pricing information, visit the Amazon Bedrock pricing page.

To get started with this feature, visit the Guardrails for Amazon Bedrock web page.

For deep-dive technical content and to learn how our Builder communities are using Amazon Bedrock in their solutions, visit our community.aws website.

— Esra

Post Syndicated from Danilo Poccia original https://aws.amazon.com/blogs/aws/agents-for-amazon-bedrock-introducing-a-simplified-creation-and-configuration-experience/

With Agents for Amazon Bedrock, applications can use generative artificial intelligence (generative AI) to run tasks across multiple systems and data sources. Starting today, these new capabilities streamline the creation and management of agents:

Quick agent creation – You can now quickly create an agent and optionally add instructions and action groups later, providing flexibility and agility for your development process.

Agent builder – All agent configurations can be operated in the new agent builder section of the console.

Simplified configuration – Action groups can use a simplified schema that just lists functions and parameters without having to provide an API schema.

Return of control –You can skip using an AWS Lambda function and return control to the application invoking the agent. In this way, the application can directly integrate with systems outside AWS or call internal endpoints hosted in any Amazon Virtual Private Cloud (Amazon VPC) without the need to integrate the required networking and security configurations with a Lambda function.

Infrastructure as code – You can use AWS CloudFormation to deploy and manage agents with the new simplified configuration, ensuring consistency and reproducibility across environments for your generative AI applications.

Let’s see how these enhancements work in practice.

Creating an agent using the new simplified console
To test the new experience, I want to build an agent that can help me reply to an email containing customer feedback. I can use generative AI, but a single invocation of a foundation model (FM) is not enough because I need to interact with other systems. To do that, I use an agent.

In the Amazon Bedrock console, I choose Agents from the navigation pane and then Create Agent. I enter a name for the agent (customer-feedback) and a description. Using the new interface, I proceed and create the agent without providing additional information at this stage.

I am now presented with the Agent builder, the place where I can access and edit the overall configuration of an agent. In the Agent resource role, I leave the default setting as Create and use a new service role so that the AWS Identity and Access Management (IAM) role assumed by the agent is automatically created for me. For the model, I select Anthropic and Claude 3 Sonnet.

In Instructions for the Agent, I provide clear and specific instructions for the task the agent has to perform. Here, I can also specify the style and tone I want the agent to use when replying. For my use case, I enter:

Help reply to customer feedback emails with a solution tailored to the customer account settings.

In Additional settings, I select Enabled for User input so that the agent can ask for additional details when it does not have enough information to respond. Then, I choose Save to update the configuration of the agent.

I now choose Add in the Action groups section. Action groups are the way agents can interact with external systems to gather more information or perform actions. I enter a name (retrieve-customer-settings) and a description for the action group:

Retrieve customer settings including customer ID.

The description is optional but, when provided, is passed to the model to help choose when to use this action group.

In Action group type, I select Define with function details so that I only need to specify functions and their parameters. The other option here (Define with API schemas) corresponds to the previous way of configuring action groups using an API schema.

Action group functions can be associated to a Lambda function call or configured to return control to the user or application invoking the agent so that they can provide a response to the function. The option to return control is useful for four main use cases:

• When it’s easier to call an API from an existing application (for example, the one invoking the agent) than building a new Lambda function with the correct authentication and network configurations as required by the API
• When the duration of the task goes beyond the maximum Lambda function timeout of 15 minutes so that I can handle the task with an application running in containers or virtual servers or use a workflow orchestration such as AWS Step Functions
• When I have time-consuming actions because, with the return of control, the agent doesn’t wait for the action to complete before proceeding to the next step, and the invoking application can run actions asynchronously in the background while the orchestration flow of the agent continues
• When I need a quick way to mock the interaction with an API during the development and testing and of an agent

In Action group invocation, I can specify the Lambda function that will be invoked when this action group is identified by the model during orchestration. I can ask the console to quickly create a new Lambda function, to select an existing Lambda function, or return control so that the user or application invoking the agent will ask for details to generate a response. I select Return Control to show how that works in the console.

I configure the first function of the action group. I enter a name (retrieve-customer-settings-from-crm) and the following description for the function:

Retrieve customer settings from CRM including customer ID using the customer email in the sender/from fields of the email.

In Parameters, I add email with Customer email as the description. This is a parameter of type String and is required by this function. I choose Add to complete the creation of the action group.

Because, for my use case, I expect many customers to have issues when logging in, I add another action group (named check-login-status) with the following description:

Check customer login status.

This time, I select the option to create a new Lambda function so that I can handle these requests in code.

For this action group, I configure a function (named check-customer-login-status-in-login-system) with the following description:

Check customer login status in login system using the customer ID from settings.

In Parameters, I add customer_id, another required parameter of type String. Then, I choose Add to complete the creation of the second action group.

When I open the configuration of this action group, I see the name of the Lambda function that has been created in my account. There, I choose View to open the Lambda function in the console.

In the Lambda console, I edit the starting code that has been provided and implement my business case:

import json

def lambda_handler(event, context):
    print(event)
    
    agent = event['agent']
    actionGroup = event['actionGroup']
    function = event['function']
    parameters = event.get('parameters', [])

    # Execute your business logic here. For more information,
    # refer to: https://docs.aws.amazon.com/bedrock/latest/userguide/agents-lambda.html
    if actionGroup == 'check-login-status' and function == 'check-customer-login-status-in-login-system':
        response = {
            "status": "unknown"
        }
        for p in parameters:
            if p['name'] == 'customer_id' and p['type'] == 'string' and p['value'] == '12345':
                response = {
                    "status": "not verified",
                    "reason": "the email address has not been verified",
                    "solution": "please verify your email address"
                }
    else:
        response = {
            "error": "Unknown action group {} or function {}".format(actionGroup, function)
        }
    
    responseBody =  {
        "TEXT": {
            "body": json.dumps(response)
        }
    }

    action_response = {
        'actionGroup': actionGroup,
        'function': function,
        'functionResponse': {
            'responseBody': responseBody
        }

    }

    dummy_function_response = {'response': action_response, 'messageVersion': event['messageVersion']}
    print("Response: {}".format(dummy_function_response))

    return dummy_function_response

I choose Deploy in the Lambda console. The function is configured with a resource-based policy that allows Amazon Bedrock to invoke the function. For this reason, I don’t need to update the IAM role used by the agent.

I am ready to test the agent. Back in the Amazon Bedrock console, with the agent selected, I look for the Test Agent section. There, I choose Prepare to prepare the agent and test it with the latest changes.

As input to the agent, I provide this sample email:

From: [email protected]

Subject: Problems logging in

Hi, when I try to log into my account, I get an error and cannot proceed further. Can you check? Thank you, Danilo

In the first step, the agent orchestration decides to use the first action group (retrieve-customer-settings) and function (retrieve-customer-settings-from-crm). This function is configured to return control, and in the console, I am asked to provide the output of the action group function. The customer email address is provided as the input parameter.

To simulate an interaction with an application, I reply with a JSON syntax and choose Submit:

{ "customer id": 12345 }

In the next step, the agent has the information required to use the second action group (check-login-status) and function (check-customer-login-status-in-login-system) to call the Lambda function. In return, the Lambda function provides this JSON payload:

{
  "status": "not verified",
  "reason": "the email address has not been verified",
  "solution": "please verify your email address"
}

Using this content, the agent can complete its task and suggest the correct solution for this customer.

I am satisfied with the result, but I want to know more about what happened under the hood. I choose Show trace where I can see the details of each step of the agent orchestration. This helps me understand the agent decisions and correct the configurations of the agent groups if they are not used as I expect.

Things to know
You can use the new simplified experience to create and manage Agents for Amazon Bedrock in the US East (N. Virginia) and US West (Oregon) AWS Regions.

You can now create an agent without having to specify an API schema or provide a Lambda function for the action groups. You just need to list the parameters that the action group needs. When invoking the agent, you can choose to return control with the details of the operation to perform so that you can handle the operation in your existing applications or if the duration is longer than the maximum Lambda function timeout.

CloudFormation support for Agents for Amazon Bedrock has been released recently and is now being updated to support the new simplified syntax.

To learn more:

• See the Agents for Amazon Bedrock section of the User Guide.
• Visit our community.aws site to find deep-dive technical content and discover how others are using Amazon Bedrock in their solutions.

— Danilo