# Build an on-premises AI Kaggle competition assistant: n8n workflow

> Started by someone opening a chat with it, and works across Qdrant. It puts a local model behind a retrieval step, so answers are drawn from documents you...

Source: https://millionrounds.com/workflow/build-an-on-premises-ai-kaggle-competition-assistant (Millionrounds; every figure on the page links to its filing or article)

[Automation library](https://millionrounds.com/workflows) · [AI assistants](https://millionrounds.com/workflows?category=AI+assistants#library)

# Build an on-premises AI Kaggle competition assistant

Started by someone opening a chat with it, and works across Qdrant. It puts a local model behind a retrieval step, so answers are drawn from documents you supply rather than from the model alone. 21 steps, 1 branch: large, and worth reading before you trust it with anything live.

Large Calls a model: a local model Retrieval

[Download free](https://millionrounds.com/login?next=%2Fapp%2Fworkflow%2Fbuild-an-on-premises-ai-kaggle-competition-assistant)

**21 steps** Large: read it before trusting it with anything live

**1** Branch: places the flow splits

**0** Loops or waits

**1** Service it connects to

## The flow

Starts it 2 Model 13 Logic 4 Service 2

The steps in the order they run, from what starts it.

- 1 **Local File Trigger** Local File trigger *Starts it*
- 2 **Recursive Character Text Splitter** Text Splitter Recursive Character Text Splitter *Model*
- 3 **Embeddings Ollama** Embeddings Ollama *Model*
- 4 **Ollama Summarizer** Lm Ollama *Model*
- 5 **When chat message received** Chat trigger *Starts it*
- 6 **Window Buffer Memory** Memory Buffer Window *Model*
- 7 **Ollama Chat Model3** Lm Chat Ollama *Model*
- 8 **Embeddings Ollama2** Embeddings Ollama *Model*
- 9 **Ollama Chat Model4** Lm Chat Ollama *Model*
- 10 **Settings** Set *Logic*
- 11 **Default Data Loader** Document Default Data Loader *Model*
- 12 **Qdrant Vector Store** Vector Store Qdrant *Model*
- 13 **Summarization Chain** Chain Summarization *Model*
- 14 **AI Agent** Agent *Model*
- 15 **Qdrant Vector Store2** Vector Store Qdrant *Model*
- 16 **Vector Store Tool** Tool Vector Store *Model*
- 17 **Import File** Read Write File *Service*
- 18 **Merge** Merge *Logic*
- 19 **Get FileType** Switch *Logic*
- 20 **Extract from TEXT** Extract From File *Service*
- 21 **Markdown** Markdown *Logic*

## How it's built

- **Starts on:** Someone opening a chat with it
- **Steps:** 21, with 1 branch
- **Size:** Large: read it before trusting it with anything live
- **Connects to:** Qdrant
- **Model:** a local model
- **Retrieval:** Answers are drawn from documents you supply, not from the model alone

## Questions

What does it need to run?

An n8n account, plus your own Qdrant account, and a key for a local model. You connect your own accounts after importing it; none are included.

What starts it?

It starts on someone opening a chat with it.

Does it use AI?

Yes. It calls a local model. Its answers draw on documents you give it, not on the model alone. Model calls are billed by the provider, not by n8n.

How long does it take to set up?

Take your time. At 21 steps it is large: read it before trusting it with anything live.

How do I import it?

Copy the workflow, open a new workflow in n8n and paste. The steps appear on the canvas, ready for your own accounts.

**Before you run it**

Nothing here has been tested by us. It is somebody's working automation, read and described from its own steps. Treat the connections it wants as a list of accounts it will touch, and read the 21 steps before you point it at anything that matters.

## Run it in n8n

- **Create your n8n account** Free to start. n8n is where the workflow runs. [Create an n8n account ↗](https://millionrounds.com/n8n)
- **Download the workflow** Free with a Millionrounds account. [Download free](https://millionrounds.com/login?next=%2Fapp%2Fworkflow%2Fbuild-an-on-premises-ai-kaggle-competition-assistant)
- **Paste it into n8n** Open a new workflow in n8n and paste. Then connect your own accounts.

Stored with the author's traces removed: canvas notes, instance identifiers, saved credential names and any live document they had left behind.

## Connects to

Each opens the other workflows that use it.

[Qdrant](https://millionrounds.com/workflows?service=Qdrant#library)

## Workflows like it

Also ai assistants, with the same services where possible, closest in size first.

- [**Gmail assistant with OpenAI** 20 steps · calls a model · Gmail, Qdrant](https://millionrounds.com/workflow/gmail-assistant-with-openai)
- [**Build a multi-agent system with Qdrant and Gmail** 23 steps · calls a model · Gmail, Google Drive](https://millionrounds.com/workflow/build-a-multi-agent-system-with-qdrant-and-gmail)
- [**Filter AI slop from your LinkedIn feed** 23 steps · calls a model · Qdrant](https://millionrounds.com/workflow/filter-ai-slop-from-your-linkedin-feed)
- [**Website chatbot with Google Drive and OpenAI** 23 steps · calls a model · Google Drive, Qdrant](https://millionrounds.com/workflow/website-chatbot-with-google-drive-and-openai)
- [**Answer restaurant voice queries with ElevenLabs and OpenAI** 18 steps · calls a model · Google Drive, Qdrant](https://millionrounds.com/workflow/answer-restaurant-voice-queries-with-elevenlabs-and-openai)
- [**Build a RAG system with Qdrant and Gemini** 24 steps · calls a model · Google Drive, Qdrant](https://millionrounds.com/workflow/build-a-rag-system-with-qdrant-and-gemini)