OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
RAG & Knowledge Systems

Serviette

Serviette is a self-hosted Python RAG server built on Pathway that indexes a document folder and serves cited answers through a web chat and REST API. Its YAML configuration selects an embedder and vector database, and it reflects document edits, additions and deletions in answers.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Serviette — OpenViST project profile in RAG & Knowledge Systems.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Serviette when you want a self-hosted, document-grounded question-answering service whose index follows changes in a watched document collection.

Who it is for

  • Teams that need question answering over an organisation's changing documents.
  • Developers setting up a self-hosted document Q&A chat or REST endpoint.
  • Operators choosing among supported embedding and vector-database integrations.

Documented uses

  • Answer questions from a maintained folder of documents and cite the sources used.
  • Expose document-grounded answers through a web chat and REST API.
  • Keep answers aligned with document edits, additions and deletions without a manual re-upload step.
Before you choose

Requirements & limits

Serviette is self-hosted and MIT-licensed. Setup requires a free Pathway license key. Provider API keys are needed for the corresponding API-based embedding or generated-answer options.

  • Python 3.10 or newer and pip 23.1 or newer are required.
  • The local embedding option uses PyTorch and has a substantial documented install; the README recommends the CPU package index for CPU use.
  • The README describes local processing with a local embedder, while retrieved passages go to the configured answer-model API; API-based embeddings also send document text to that provider.
  • The service is question-by-question and has no built-in conversation memory or user accounts.
  • The repository describes eight vector-database integrations, while the supplied metadata names ChromaDB, DuckDB, Milvus, MongoDB, pgvector and Pinecone.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Configure and start Serviette over your own documents

Generate a configuration and start the indexer and server together for a document-backed chat and REST API.

Before you start

  • Python 3.10 or newer and pip 23.1 or newer
  • A free Pathway license key set in PATHWAY_LICENSE_KEY
  • A document folder and a chosen embedder and vector database
  • For local embeddings, serviette[local] and the documented PyTorch CPU package index; for OpenAI embeddings, an OPENAI_API_KEY
  • An OpenAI API key for generated answer text; the README also describes a keyless mode that quotes retrieved passages

Steps

  1. Install Serviette with the documented local-embedding option or install the base package if using OpenAI embeddings.

  2. Set PATHWAY_LICENSE_KEY and any required provider key in the environment.

  3. Run serviette wizard to generate a configuration file.

  4. Run serviette up --config config.yaml to start the indexer and server together.

Expected outcome

A server exposing the web chat and versioned REST API, grounded in the configured documents and able to cite sources.

Guide limitations

  • The guide covers setup and startup; configuration-specific document-source and vector-database settings depend on the selected options.
  • The service answers one question at a time and does not include conversation memory or user accounts.
Documentation 1 ↗Documentation 2 ↗
COMPARE OPTIONS

Compare options for Vector search

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.

ProjectDocumented purposePlatformAccess and requirements
ServietteAnswer questions from a maintained folder of documents and cite the sources used.Linux, Self-hosted, ServerServiette is self-hosted and MIT-licensed. Setup requires a free Pathway license key. Provider API keys are needed for the corresponding API-based embedding or generated-answer options.
Chatbot with PDF for Semantic SearchPrepare document content as embeddings and save it to a supported vector database.WebThe repository is identified as open source and publicly visible on GitHub.
AI LLM PlaygroundAsk questions about PDF documents through an assistant interface.Web
Evidence and transparency

Sources & review

Official repository ↗Guide documentation ↗
Source record details
Source availability
Active
Sources checked
20 October 2026
Original website title
pathwaycom/serviette · Official homepage title · retained in its source language
How we review sourcesSuggest a correction