Build Agent Pipelines in AgentCreator

These steps focus on describing the generic process workflow for creating an Agent in SnapLogic.

Make sure your LLM Snap accounts are configured. Refer to the following for LLM configuration details:
- Amazon Bedrock LLM accounts
- Configure the OpenAI LLM accountOpenAI LLM accounts
- Azure OpenAI LLM accounts
- Google GenAI LLM accounts
Configure Snap accounts for data and application endpoints available for tool calling.
Set up your Snaplex to validate and run your pipelines.

Plan your agent pipeline design.
1. Objective - defining a specific goal for your agent is crucial as a first step.
2. Preparation - Collect data and prepare a real-world context for the types of end-user questions the Agent is designed to answer.
3. Design - pipeline design includes the layers of pipelines and informs the hierarchy of Agents.
4. Tools - tool planning might be the most intensive phase because in addition to selecting the tools, attaining access and defining the use of the tool are crucial.
5. Data sources - identifying data sources and getting access to them are a necessity for a successful Agent. This applies to RAG pipelines especially.
In Designer, create the worker pipeline by dragging the required Snaps onto the canvas and connecting them.
Tip:
You can build one worker pipeline and add all of your tools in this pipeline. The advantage is that you can work in one Designer canvas to design each tool, making the following activities more efficient:
- Tool validation
- Testing response from different LLM models
- Snap account usage
You can have multiple unconnected pipelines on one canvas. Use the Disable Snap functionality to turn off pipeline segments from validating and executing.
1. For each tool pipeline branch, start with the Function Generator Snap to create a function definition for the tools to be called.
  You can connect multiple Function Generator Snaps to call different tools.
2. Add the Message Appender to pass the message payload (system and user prompts), and tools payload (function definitions) to the Tool Calling Snap
3. Add the tool pipeline branch for your Function calling pipeline to use.
4. Use the HTTP Client Snap to connect to an online web applications. It sends HTTP requests to a server and receive corresponding HTTP responses.
5. Insert a Mapper Snap to define data mappings between systems.
6. Make sure that the starting Snap of your tool and worker pipelines have an open view.
  Tip: Use a JSON Generator Snap to mock data input for testing purposes. It can be disconnected later and disabled.
When your tool testing is complete, and you've validated all of your tool pipelines, you need to extract each tool pipeline branch into its own Designer canvas and save it as a separate pipeline.
Go to Designer and create the following pipeline:

From the Snap Catalog, add the following Snaps to the canvas in order:
1. JSON Generator Snap
  
  Use the JSON editor for the initial prompt. This creates the JSON document that is sent through the pipeline. The JSON document is for mock data.
  
  Note: Once the pipeline is in production, you should disable or remove the JSON Generator Snap. Typically, Agent pipelines are put into production as Triggered Tasks powering a Agent UI. The open input for the data source would trigger the pipeline.
2. Prompt Generator Snap
  1. Click Prompt Composer and enter your prompt Prompt.
  2. Click System Prompt and define the role.
  3. Add another Prompt Generator Snap and enter the user prompt.
3. Pipe Loop
  Configure the Pipe Loop to call the Worker Agent pipelines. This might take some fine-tuning depending on your requirements. To learn about enhanced controls in the Pipe Loop Snap, refer to A Guide to the Enhanced PipeLoop.
Test and evaluate the worker pipeline against the driver pipelines.

Tip: You can mock LLM prompt input by adding a JSON Generator Snap to your pipeline.
Deploy your Agent to production and monitor its usage.
These activities include:
- Create a Triggered Task for the pipeline.
- Move the Task, pipelines, and associated assets (such as an accounts file or expression library) to a production Environment (Org).
- Assign the pipeline a Snaplex in your Production environment.
- View execution statistics in Monitor.
Activities outside of SnapLogic would be:
- Create an interface for your end users to interact with the Agent.
- Use a 3rd-party monitoring tool to assess quality of responses and LLM costs.

Validate and test your pipelines in a sandbox environment. Use Agent Visualizer to debug pipeline issues.
This Snap offers a graphical representation of the various layers of pipelines in your agent. It also displayed the prompt input and output across the pipelines.