Chat with ALL your Documents with Regolo + Elysia

Francesco Massa∙14 September 2025

Introduction 🤖📚

Would you like to have a chatbot with ALL your documents in a private database? 🔒
Now with Regolo + Elysia, you can create a true second brain 🧠 powered by Agentic RAG, keeping your knowledge secure while scaling sustainably.

In this article, we’ll explore how to integrate Regolo.ai models into Elysia in a fun, practical way. 🚀

What You Will Learn 🎯

This integration allows you to:
✅ Use Regolo.ai models on your Weaviate datasets
✅ Configure Base Model and Complex Model directly in Elysia’s UI
✅ Create your own FREE databases with Weaviate 🗄️

Never lose information again thanks to your decision-making agent capable of reasoning and tool calling 🛠️

What is Elysia 🌐

Elysia is a framework by Weaviate that lets you build AI agents based on decision trees 🌳.
Its superpower? Information Retrieval with Agentic RAG ⚡ your memory grows with your data, and the only limit is how much you want to feed it.

With Elysia you can:

🌟 Tap into a decision-tree agentic architecture, where the AI dynamically chooses tools (query, summarize, aggregate, custom) based on goals, context, and history all transparent and tweakable.

📊 Use dynamic data displays: tables, charts, doc cards, ecommerce views, conversations… Elysia picks the best format for your data automatically.

🧠 Benefit from built-in data awareness: your collections get summarized, metadata generated, and search suggestions improved for smarter retrieval.

🔁 Rely on a feedback-driven learning loop: thumbs-up responses train the agent over time, letting smaller models perform like giants.

📄 Apply chunk-on-demand document handling: no more wasteful pre-chunking; Elysia splits docs only when needed, preserving context.

Prerequisites ⚙️

🐍 Python 3.12+ and Git installed
🔑 A Regolo.ai API key

Quick Start Guide

1) Get a Regolo.ai API key 🔑

Generate an API key from Regolo.ai and keep it handy.

2) Install Elysia🌐

Open github and follow the elysia installation instructions

3) Start Elysia ▶️

elysia start

Browse on the port that just opened

5) Open Settings ⚙️

On the Left side menu navigate to Settings:

6) Run Weaviate Locally for Your Regolo + Elysia Environment 🔗

You can install Weaviate on your own machine or server to manage your document embeddings and vector search without relying on any external services.
Here’s a quick guide to running Weaviate self-hosted.

Quickstart: Run Weaviate Locally with Docker Compose

1. In your project directory, create a file named docker-compose.yml with the following content:

services:
  weaviate:
    image: cr.weaviate.io/semitechnologies/weaviate:1.33.0
    ports:
      - "8082:8080"
      - "50051:50051"
    environment:
      QUERY_DEFAULTS_LIMIT: 25
      AUTHENTICATION_ANONYMOUS_ACCESS_ENABLED: 'true'
      PERSISTENCE_DATA_PATH: '/var/lib/weaviate'
      ENABLE_MODULES: 'text2vec-ollama,generative-ollama'
      CLUSTER_HOSTNAME: 'node1'
    volumes:
      - weaviate_data:/var/lib/weaviate
    restart: on-failure:0

volumes:
  weaviate_data:

Code language: YAML (yaml)

2. Start Weaviate Using Docker Compose:

docker-compose up -d

3. Once running, your Weaviate instance will be available at http://localhost:8082.
All vector data will be stored persistently inside the Docker volume weaviate_data.

4. Test the connection:

import weaviate

client = weaviate.connect_to_custom(
    http_host="localhost",
    http_port=8082,
    http_secure=False,
    grpc_host="localhost",
    grpc_port=50051,
    grpc_secure=False
)

try:
    print(client.is_ready())  # Should return True if server is active
finally:
    client.close()Code language: Python (python)

Further Configuration and Options

You can enable additional modules such as text2vec-openai, qna-transformers, or generative-ollama by editing the ENABLE_MODULES environment variable.
For production or distributed setups, Weaviate also supports Kubernetes deployments.
Data persistence is handled automatically through the /var/lib/weaviate directory.
Official SDKs are also available for JavaScript, Go, and Java, making integration with Elysia and Regolo seamless across different environments.

For more technical details, refer to the official Weaviate documentation:
Quickstart Local Installation ·
Client Connections ·
Deployment Guides

7) Upload your Documents 🤖

We created a personalized document embedder based on Regolo’s Qwen3-Embedding-8B, it directly uploads the extracted data in your 8082 weaviate database, ready to chat.
You can find it and use it on github (copy it on a dedicated folder):

git clone https://github.com/regolo-ai/regolo-docs-embedderCode language: PHP (php)

8) Configure the Database on the Platform 🤖

Paste the DB Settings into Elysia (Use the same settings for Weaviate Cluster and Elysia Storage)

9) Agent Setup 🤖

You Can Modify it as you want, it depends on you Elysia final goal

10) Base Model – Regolo Configuration 🧩

In this section you have to use “openai” provider and write the name of a custom Regolo model (example : Qwen3-8B )

11) Complex Model – Regolo Configuration 🧩

As in the previous step, insert your favourite Regolo Big Model

# our advice is to use one of
Llama-3.3-70B-Instruct
gpt-oss-120b
deepseek-r1-70b) Code language: CSS (css)

with “openai” Provider

12) Base URL 🌍

This step is crucial, insert the Regolo Base Url:

https://api.regolo.ai/v1Code language: JavaScript (javascript)

You should get something like this:

13) Api Key 🔑

In the Api Keys Section insert both “regolo_api_key” and “openai_api_key” (same value) like this:

🎉Finished ! 🎉

That’s it, your Elysia + Regolo setup is live!
Now you can chat with your private data, explore your Weaviate collections, and let Regolo power your second brain whenever you want. ⚡

Tutorial