Practical overview
When it fits
Use Neumai when you need to assemble or manage the ingestion, embedding and vector-storage stages that supply contextual data to an LLM-based RAG system.
Who it is for
- Developers building retrieval-augmented generation systems from their own documents or database data.
- Teams that need to connect data sources, embedding services and vector stores in a pipeline.
Documented uses
- Ingest website content or PostgreSQL data and prepare it for vector search.
- Choose content fields for embedding and metadata fields for retrieval context.
- Generate embeddings and write them to a vector database.
- Search a pipeline for relevant results and access their metadata.
- Keep embeddings synchronized as source data changes.
Requirements & limits
The README documents local development through the neumai Python package, a cloud path that uses a Neumai API key, and self-hosting by contacting the team.
- The shown website pipeline uses OpenAI embeddings and Weaviate, so it requires credentials and configuration for both services.
- The PostgreSQL example uses a connection string and query, JSON loading, and field selection for embedded content and metadata.
- Local development, cloud publishing and self-hosting are separate paths; the self-hosting instructions point to contacting the team and a sample backend architecture.
- The repository lists Python as its primary language and Apache-2.0 as its license.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Build and query a website-to-Weaviate pipeline
Process a website page into vectors in Weaviate and search the pipeline for relevant content.
Before you start
- Install the neumai Python package.
- An OpenAI API key.
- A Weaviate URL, API key and class name.
Steps
Create a WebsiteConnector for the page and select URL metadata.
Wrap the connector in a SourceConnector with an HTMLLoader and RecursiveChunker.
Configure OpenAIEmbed and WeaviateSink with their respective credentials and settings.
Create a Pipeline with the source, embed connector and sink, then call pipeline.run().
Call pipeline.search() with a query and result count; the example prints result metadata.
Expected outcome
A pipeline that embeds website content, stores vectors in Weaviate and returns search results with metadata.
Guide limitations
- The example demonstrates one website page, OpenAI embeddings and Weaviate; it does not document other connector configurations here.
- The README provides example code, not a guarantee that a particular environment or deployment will work unchanged.
Compare options for Retrieval-augmented applications
These projects document a common task and share a project role. This is a source-based comparison, not an execution benchmark or a guarantee of interchangeability.
| Project | Documented purpose | Platform | Access and requirements |
|---|---|---|---|
| Neumai | Ingest website content or PostgreSQL data and prepare it for vector search. | Self-hosted, Server | The README documents local development through the neumai Python package, a cloud path that uses a Neumai API key, and self-hosting by contacting the team. |
| Langsearch | Crawl website or local data for a knowledge base. | CLI, Self-hosted, Server | |
| Hybrid Search Rag | Combine document retrieval and language-model generation in a configurable workflow. |
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 15 October 2026
- Original website title
- NeumTry/NeumAI · Official homepage title · retained in its source language