Built by Metorial, the integration platform for agentic AI.

Learn More

Tools

Create Chatbot

Create a new chatbot in VectorShift backed by a pipeline. Chatbots can be used for customer support, onboarding, lead collection, and advisory workflows.

List Knowledge Bases

List all knowledge bases in the VectorShift account. Optionally include shared knowledge bases and retrieve full details.

Add Data to Knowledge Base

Add data to an existing knowledge base. Supports adding content via URL (with optional recursive crawling) or plain text. The data will be processed, chunked, and indexed for semantic search.

Delete Chatbot

Permanently delete a chatbot from VectorShift. This action cannot be undone.

Create Knowledge Base

Create a new knowledge base in VectorShift. Knowledge bases store and index documents for semantic search within AI workflows. Configure document processing options such as chunk size, overlap, and processing implementation.

Query Knowledge Base

Perform semantic search queries against a knowledge base. Returns the most relevant document chunks matching your queries. Supports multiple queries in a single request with optional context, result limits, reranking, and query transformation.

List Pipelines

List all pipelines in the VectorShift account. Optionally include shared pipelines and retrieve full pipeline details.

Delete Pipeline

Permanently delete a pipeline from VectorShift. This action cannot be undone.

Run Chatbot

Send a message to a VectorShift chatbot and receive a response. Supports multi-turn conversations by passing a conversation ID. Use for customer support, onboarding, lead collection, and advisory use cases.

Bulk Run Pipeline

Execute a VectorShift pipeline multiple times in a single request. Each run receives its own set of input key-value pairs. Useful for batch processing scenarios where the same pipeline needs to process multiple datasets.

Delete Transformation

Permanently delete a transformation from VectorShift. This action cannot be undone.

Run Transformation

Execute a transformation (arbitrary code) in VectorShift's isolated execution environment. Provide input key-value pairs matching the transformation's schema and receive computed results.

List Chatbots

List all chatbots in the VectorShift account. Optionally include shared chatbots and retrieve full chatbot details.

Get Knowledge Base

Fetch details of a specific knowledge base by its ID or name. Returns the full knowledge base configuration and metadata.

Run Pipeline

Execute a VectorShift pipeline with the provided inputs. Pipelines are AI workflows that process data through connected components. Provide input key-value pairs matching the pipeline's expected inputs and receive the computed outputs.

Manage Knowledge Base Documents

List or delete documents within a knowledge base. Use to browse indexed documents, filter by title, or remove specific documents from the knowledge base.

Get Transformation

Fetch details of a specific transformation by its ID or name. Returns the full transformation configuration including function code, input/output schemas, and metadata.

List Transformations

List all transformations in the VectorShift account. Optionally include shared transformations and retrieve full details.

Terminate Chatbot Session

Terminate an active chatbot conversation session. Use this to end long-running or stuck chatbot sessions.

Control Pipeline Run

Control an active pipeline run by terminating, pausing, or resuming it. Use this to manage long-running or stuck pipeline executions.

Get Chatbot

Fetch details of a specific chatbot by its ID or name. Returns the full chatbot configuration, deployment status, and metadata.

Delete Knowledge Base

Permanently delete a knowledge base and all its indexed documents from VectorShift. This action cannot be undone.

Get Pipeline

Fetch details of a specific pipeline by its ID or name. Returns the full pipeline configuration and metadata.

Create Transformation

Create a new transformation in VectorShift. Transformations allow you to run arbitrary code in an isolated execution environment. Define the function code, input/output schemas, and metadata.