Create an agent who knows the area using Amazon Bedrock Agents and four APIs
This post is written on Vikram Gunundoti and Nate Folkert from Fourquare.
Making your preferences is important to developing memorable experiences. Even if it recommends a complete movie or suggesting a new restaurant, TAALINGINGS of each person's options can make all the difference. But when it comes to eating and jobs, there is more to consider one's taste. The location and weather also plays an important role in creating our decisions. Imagine planning a day: Forn afternoon, leisure picnic park might be a good thing, but if it poured rain, the cool DOOR café would advertise. The challenge, creating an agent that can include these items – the weather, the weather and preference of a person's own – to provide genuine actions.
To also deal with this challenge, we can include Amazon Bedrock Agents and four appropriate APIs. In this post, we show how to use the area that is coming to bring your users' answers.
Amazon Bedrock Agents
Amazon Bedrock makes straight straight to create and measure AI apps. It provides access to the various main in districts (FMS) from the leading companies such as AI21 Labs, Anthropic, Retho, Meta, Ai Steaability, Amazon, all in one API. This means that you do not need to manage infrastructure, because it is also done to meet regular AWS security services, privacy, and AI. You can view models, make personal data, and create apps without writing a complex code.
Amazon Bedrock agents There is an Amazon Bedrock feature that allows you to create Agents Ai Aenterous. These agents can understand user requests, break the steps, and complete the functions by connecting to your company's API and data resources. For example, they can use such procedures as the processing of insurance claims or manage the inventory, making them business activities. They treat quick engineering, memory, and protection automatically, so you can put yourself quickly without treating infrastructure.
Four APIS areas
Foulquare's Places API brings a straight local environmental intelligence of applications that need to raise awareness. Designed on top of Open Global Data Areas with 100 million scales interesting 1,500 categories, the Areas Apis Transform geographic links the context of the useful business.
Geotagging API is accurately setting the GPS coordinating in a specific logical area, enabling apps to quickly reach that the user is in a local coffee shop, inside the Macy-in-store shop. Location Search and Apis Data Change application that apps get locations by providing sensitive sorting skills in this simple search skills. Engineers have not sorted areas only in certain phases (finding restaurants, or fascinating tourists), using the current construction issues, as well as the current performance status), and the performance of the performance in accordance with the effects of the truth. Each restored area comes developed with Kingdom qualities, including photos, reviews, quality ratings, and actual preferences.
When combined with Amazon Bedrock Agents, Foursquare Places Apis enables the creation of informal applications in the appropriate, relevant, effective user.
Looking for everything
To show energy that adds an Amazon Bedrock Agents, we create a simple construction of Amazon Bedrock Agent with APIs with four areas and the weather apts and the weather apts. By combining these skills, we can create a different customized user experience in context. The next painting shows how we testified.
In a solution period, the user interacts with an agent with a measuring web display. The web app uses a logic app that costs Amazon Bedrock Agent in the cloud. Agel knows local tools and weather tools although there is a location within the app. When the Agent's Tools, Reference Office Officer of the app provided, providing a tool and provides feedback from second reduction tool. In addition to the tools, an agent has fundamental orders of what kind of personality is to have and which types of behavior should support.
Let us examine the example of short communication with an agent when we ask if there is a park near the recommended restaurant near the food park.
The following screen shows the initial partnership with agent, getting a park near the four Apises requested by the agent.
In this example, you can see an agent of sending out the Civil Events to the user informing them of the world's actions (requesting a model, requesting a tool, and thinking).
This next screen shows a restaurant recommended by the four Apis near the park.
In this example, the agent urges the APIs' based APIs, and Sislowlit Ui links outgoing from Foursquare to the map.
In the following section, we explain how to create an agent in your account and start.
Requirements
To use this solution, you must have an AWS account that has the necessary permissions.
You will also need the API API's API of API of Foursquare API to allow the APA agent to reach the end of four APIs. If you don't already, follow the instructions to four documents – Manage your API service keys to create one. You will need to sign in to your engineer's Foursquare account or create one if you do not have one (creating a basic account and includes credit for your project). Be sure to copy the key to Service API key in the construction as you will be able to see you again.
Form an agent
The Foursquare Agent Code is available as an open source in the next GitTub area. Complete the following steps to upgrade an agent in your local folder from the Source:
- Clone is a local folder site.
- Set the environmental variables for your Foursquare API Token:
export FOURSQUARE_SERVICE_TOKEN=
- Set the environmental variables with your AWS guarantees:
export AWS_ACCESS_KEY_ID=
export AWS_SECRET_ACCESS_KEY=
- Enter the requirements:
pip install requirements.txt
- Start UI broadcasts:
streamlit run agent_ui.py
The best habits
If you make a agent, we recommend starting with the test data. Consider the installation possible and what is acceptable results. Use these sample discussion to check the agent whenever changes. In addition, Amazon Bedrock Agents allows you to prepare Guardrails to protect the malicious installation or types of discussion that you would not want to use your user experience. We recommend that any charges for production records your couple with appropriate gangs. To learn more, see the Amazon Bedrock Guardrails.
Clean
When you are finished using the solution, remove any services you have made to avoid ongoing crimes.
Store
Agents provide a function of automatic function on behalf of your customers, either by using a conversation interface or other input interface. Integrating Automation may also have an API API API from Foursquare, You can create powerful UI and fanatic Servers for new performance levels. With Amazon Bedrock Agents, you can create a focal solution that allows you to use the powerful base models of Amazon Bedrock to drive this experience.
Try your case solution, and share your answer to a comment.
About the authors
John Baker Is the primary AWS SDE, where it works in Amazon Bedrock and Amazon Bedrock Agents. He has been with Amazon for more than 10 years and work across the AWS, Alexa, and Commazon.com. In his spare time, John is enjoying skiing and other external activities in all Pacific Northwest.
Mark Roy Has the Customer Awspaper artist, helping customers design and create solutions for AIs. His focus since the beginning of 2023 has been lending to efforts to build a Dazon Bedrock launch, AI evolutionary AI from AWS. Mark's work includes a comprehensive range of cases, primary enthusiasm in making AI, agents, and measuring ML across the business. He has assisted in the insurance companies, financial services, media and entertainment, health care, services, and production. Before joining the AWS, Mark was an artist of property, engineer, and technical technology for more than 25 years, including 19 years of financial resources. Mark holds six AWS certificates, including a special ML certificate.
Vikram Gunundete It is currently working as a Foursquare specialist, where he led a technical strategy, making decisions, and a plan for geospatial company platform. Before joining the Foursquare, Vikram caught a central Amazon engineer, where he made his marker as an engineer that invented a group of Amazon Alexa.
Nate Folkert He is the engineer of the top workers with a Foursquare, where he has seen right there near he looked at the SOho coffee shop 14 years ago. He built a server API for Sesharm and helps in special projects. Outside, do you like to explore the world (gearm, ofc, so isn't it without work?) And is currently worried about finding all the IRL TV filters Overdoing too much
