This n8n workflow is designed to streamline the analysis and logging of packaging waste data through an automated process involving webhooks, natural language processing, AI embeddings, and data storage. It is ideal for organizations seeking to collect input data related to packaging waste, analyze it using AI models, and log the results for future reference.
The process begins when the workflow is triggered via an HTTP POST webhook at the endpoint ‘packaging_waste_calculator’. The incoming data, received by the webhook, is processed using a text splitter node which divides the text into manageable chunks of 400 characters with 40 characters of overlap.
Next, these chunks are fed into OpenAI’s embeddings model to generate vector representations, which are then stored into a Redis vector database for efficient retrieval. The workflow allows querying of this stored data to find relevant information based on similarity.
A language model chat with Anthropic’s API takes these relevant data points and processes them through an agent node, which is prompted to define or summarize the data. The output from this step is a coherent analysis or explanation.
Finally, the results are logged into a Google Sheet under a specific sheet called ‘Log’ for record-keeping and further analysis. This automated pipeline is useful for organizations tracking packaging waste, enabling quick data analysis, and maintaining logs with minimal manual intervention.
Reviews
There are no reviews yet.