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

Modular Rag

Modular Rag is an open-source Java sample project that demonstrates modular retrieval-augmented generation patterns using Spring AI and Ollama, with optional Tavily Search and development services for document processing and observability.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Modular Rag — 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 it when you want a Java reference for exploring RAG components and a local development workflow, rather than a hosted service for managing a knowledge base.

Who it is for

  • Java developers exploring retrieval-augmented generation architecture.
  • Developers comparing local Ollama workflows and optional web search in a RAG application.
  • Teams examining document-processing and observability services in a development setup.

Documented uses

  • Study a modular Spring AI RAG implementation.
  • Run a sample application against an Ollama inference server.
  • Experiment with optional Tavily Search in the application.
  • Inspect development-service integrations for PostgreSQL, document processing, and observability.
Before you choose

Requirements & limits

The repository is public and licensed under Apache-2.0. Running the sample requires Java 25 and Podman or Docker; Tavily Search requires an API key when used.

  • Ollama can be run locally or provisioned through Arconia Dev Services.
  • Tavily Search is optional and configured through the TAVILY_SEARCH_API_KEY environment variable.
  • The documented development services include PostgreSQL, Docling Serve, and Grafana LGTM or Arize Phoenix.
  • Grafana's dashboard URL is shown in application logs; telemetry is exported over OTLP using HTTP/Protobuf.

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

Start the sample with a local Ollama server

Run the application using an Ollama inference server already running locally.

Before you start

  • Java 25
  • Podman or Docker
  • A locally running Ollama inference server

Steps

  1. Make sure the local Ollama server is running.

  2. From the project, run ./gradlew bootRun or use the documented arconia dev command.

  3. Open the application at http://localhost:8080.

Expected outcome

The sample application is available locally; Spring AI pulls required Ollama models at startup if they are not already present.

Guide limitations

  • The steps describe local application startup, not a production deployment.
  • Tavily Search is optional and requires a separately configured API key.
Documentation 1 ↗
Source-based practical guide

Enable the Ollama development service

Start the sample with Ollama provisioned through Arconia Dev Services.

Before you start

  • Java 25
  • Podman or Docker
  • Arconia Dev Services and Testcontainers setup

Steps

  1. From the project, run ./gradlew bootRun -Darconia.dev.services.ollama.enabled=true.

  2. Open the application at http://localhost:8080.

Expected outcome

The sample starts with the Ollama Dev Service enabled; Spring AI pulls required models if they are not already available.

Guide limitations

  • This route depends on the documented Arconia Dev Services and Testcontainers setup.
  • Tavily Search remains optional and is not enabled by this command.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
16 October 2026
Original website title
ThomasVitale/modular-rag · Official homepage title · retained in its source language
How we review sourcesSuggest a correction