LLM App

Pathway's LLM (Large Language Model) Apps allow you to quickly put in production AI applications which use the most up-to-date knowledge available in your data sources. You can directly run a 24/7 service to answer natural language queries about an ever-changing private document knowledge base, or run an LLM-powered data transformation pipeline on a data stream.

The Python application examples provided in this repo are ready-to-use. They can be run as Docker containers, and expose an HTTP API to the frontend. To allow quick testing and demos, most app examples also include an optional Streamlit UI which connects to this API. The apps rely on the Pathway framework for data source synchronization, for serving API requests, and for all low-latency data processing. The apps connect to document data sources on S3, Google Drive, Sharepoint, etc. with no infrastructure dependencies (such as a vector database) that would need a separate setup.

Quick links - 👀 Why use Pathway LLM Apps? 🚀 Watch it in action 📚 How it works 🌟 Application examples 🏁 Get Started 💼 Showcases 🛠️ Troubleshooting 👥 Contributing ⚙️ Hosted Version 💡 Need help?

Why use Pathway LLM Apps?

1. Simplicity - Simplify your AI pipeline by consolidating capabilities into one platform. No need to integrate and maintain separate modules for your Gen AI app: Vector Database (e.g. Pinecone/Weaviate/Qdrant) + Cache (e.g. Redis) + API Framework (e.g. Fast API).
2. Real-time data syncing - Sync both structured and unstructured data from diverse sources, enabling real-time Retrieval Augmented Generation (RAG).
3. Easy alert setup - Configure alerts for key business events with simple configurations. Ask a question, and get updated when new info is available.
4. Scalability - Handle heavy data loads and usage without degradation in performance. Metrics help track usage and scalability. Learn more about the performance of the underlying Pathway data processing framework.
5. Monitoring - Provide visibility into model behavior via monitoring, tracing errors, anomaly detection, and replay for debugging. Helps with response quality.
6. Security - Designed for Enterprise, with capabilities like Personally Identifiable Information (PII) detection, content moderation, permissions, and version control. Pathway apps can run in your private cloud with local LLMs.
7. Unification - Cover multiple aspects of your choice with a unified application logic: back-end, embedding, retrieval, LLM tech stack.

Watch it in action

Effortlessly extract and organize unstructured data from PDFs, docs, and more into SQL tables - in real-time.

Analysis of live documents streams.

(See: unstructured-to-sql app example.)

Automated real-time knowledge mining and alerting.

Monitor streams of changing documents, get real-time alerts when answers change.

Using incremental vector search, only the most relevant context is automatically passed into the LLM for analysis, minimizing token use - even when thousands of documents change every minute. This is real-time RAG taken to a new level 😊.

For the code, see the drive_alert app example. You can find more details in a blog post on alerting with LLM-App.

How it works

The default contextful app example launches an application that connects to a source folder with documents, stored in AWS S3 or locally on your computer. The app is always in sync with updates to your documents, building in real-time a "vector index" using the Pathway package. It waits for user queries that come as HTTP REST requests, then uses the index to find relevant documents and responds using OpenAI API or Hugging Face in natural language. This way, it provides answers that are always best on the freshest and most accurate real-time data.

This application template can also be combined with streams of fresh data, such as news feeds or status reports, either through REST or a technology like Kafka. It can also be combined with extra static data sources and user-specific contexts, to provide more relevant answers and reduce LLM hallucination.

Read more about the implementation details and how to extend this application in our blog article.

Instructional videos

▶️ Building an LLM Application without a vector database - by Jan Chorowski

▶️ Let's build a real-world LLM app in 11 minutes - by Pau Labarta Bajo

Advanced Features

LLM Apps built with Pathway can also include the following capabilities:

• Local Machine Learning models - Pathway LLM Apps can run with local LLMs and embedding models, without making API calls outside of the User's Organization.
• Multiple live data sources - Pathway LLM Apps can connect to live data sources of diverse types (news feeds, APIs, data streams in Kafka, and others),
• Extensible enterprise logic - user permissions, user session handling, and a data security layer can all be embedded in your application logic by integrating with your enterprise SSO, AD Domains, LDAP, etc.
• Live knowledge graphs - the Pathway framework enables concept mining, organizing data and metadata as knowledge graphs, and knowledge-graph-based indexes, kept in sync with live data sources.

To learn more about advanced features see: Features for Organizations, or reach out to the Pathway team.

Application Examples

Pick one that is closest to your needs.

Example app (template)	Description
demo-question-answering	The question-answering pipeline that uses the GPT model of choice to provide answers to the queries about a set of documents. You can also try it on the Pathway Hosted Pipelines website.
demo-document-indexing	The real-time document indexing pipeline that provides the monitoring of several kinds of data sources and health-check endpoints. It is available on the Pathway Hosted Pipelines website.
contextless	This simple example calls OpenAI ChatGPT API but does not use an index when processing queries. It relies solely on the given user query. We recommend it to start your Pathway LLM journey.
contextful	This default example of the app will index the jsonlines documents located in the data/pathway-docs directory. These indexed documents are then taken into account when processing queries.
contextful-s3	This example operates similarly to the contextful mode. The main difference is that the documents are stored and indexed from an S3 bucket, allowing the handling of a larger volume of documents. This can be more suitable for production environments.
unstructured	Process unstructured documents such as PDF, HTML, DOCX, PPTX, and more. Visit unstructured-io for the full list of supported formats.
local	This example runs the application using Huggingface Transformers, which eliminates the need for the data to leave the machine. It provides a convenient way to use state-of-the-art NLP models locally.
unstructured-to-sql	This example extracts the data from unstructured files and stores it into a PostgreSQL table. It also transforms the user query into an SQL query which is then executed on the PostgreSQL table.
alert	Ask questions, get alerted whenever response changes. Pathway is always listening for changes, whenever new relevant information is added to the stream (local files in this example), LLM decides if there is a substantial difference in response and notifies the user with a Slack message.
drive-alert	The alert example on steroids. Whenever relevant information on Google Docs is modified or added, get real-time alerts via Slack. See the tutorial.
contextful-geometric	The contextful example, which optimises use of tokens in queries. It asks the same questions
with increasing number of documents given as a context in the question, until ChatGPT finds an answer.

Get Started

To run the demo-document-indexing vector indexing pipeline and UI please follow instructions under examples/pipelines/demo-document-indexing/README.md.

To run the demo-question-answering question answering pipeline please follow instructions under examples/pipelines/demo-question-answering/README.md.

For all other demos follow the steps below.

Prerequisites

1. Make sure that Python 3.10 or above installed on your machine.
2. Download and Install Pip to manage project packages.
3. [Optional if you use OpenAI models]. Create an OpenAI account and generate a new API Key: To access the OpenAI API, you will need to create an API Key. You can do this by logging into the OpenAI website and navigating to the API Key management page.
4. [Important if you use Windows OS]. The examples only support Unix-like systems (such as Linux, macOS, and BSD). If you are a Windows user, we highly recommend leveraging Windows Subsystem for Linux (WSL) or Dockerize the app to run as a container.
5. [Optional if you use Docker to run samples]. Download and install docker.

Now, follow the steps to install and get started with one of the provided examples. You can pick any example that you find interesting - if not sure, pick contextful.

Alternatively, you can also take a look at the application showcases.

Step 1: Clone the repository

This is done with the git clone command followed by the URL of the repository:

git clone https://github.com/pathwaycom/llm-app.git

Next, navigate to the repository:

cd llm-app

Step 2: Set environment variables

Create an .env file in the root directory and add the following environment variables, adjusting their values according to your specific requirements and setup.

Environment Variable	Description
APP_VARIANT	Determines which pipeline to run in your application. Available modes are [contextful, contextful-s3, contextless, local, unstructured-to-sql, alert, drive-alert]. By default, the mode is set to contextful.
PATHWAY_REST_CONNECTOR_HOST	Specifies the host IP for the REST connector in Pathway. For the dockerized version, set it to 0.0.0.0 Natively, you can use 127.0.0.1
PATHWAY_REST_CONNECTOR_PORT	Specifies the port number on which the REST connector service of the Pathway should listen. Here, it is set to 8080.
OPENAI_API_KEY	The API token for accessing OpenAI services. If you are not running the local version, please remember to replace it with your API token, which you can generate from your account on openai.com.
PATHWAY_PERSISTENT_STORAGE	Specifies the directory where the cache is stored. You could use /tmpcache.

For example:

APP_VARIANT=contextful
PATHWAY_REST_CONNECTOR_HOST=0.0.0.0
PATHWAY_REST_CONNECTOR_PORT=8080
OPENAI_API_KEY=<Your Token>
PATHWAY_PERSISTENT_STORAGE=/tmp/cache

Step 3: Build and run the app

You can install and run your chosen LLM App example in two different ways.

Using Docker

Docker is a tool designed to make it easier to create, deploy, and run applications by using containers. Here is how to use Docker to build and run the LLM App:

docker compose run --build --rm -p 8080:8080 llm-app-examples

If you have set a different port in PATHWAY_REST_CONNECTOR_PORT, replace the second 8080 with this port in the command above.

When the process is complete, the App will be up and running inside a Docker container and accessible at 0.0.0.0:8080. From there, you can proceed to the "Usage" section of the documentation for information on how to interact with the application.

Native Approach

• Install poetry:

  pip install poetry

• Install llm_app and dependencies:

  poetry install --with examples --extras local

  You can omit --extras local part if you're not going to run local example.

• Run the examples: You can start the example with the command:

  poetry run ./run_examples.py contextful

Step 4: Start to use it

1. Send REST queries (in a separate terminal window): These are examples of how to interact with the application once it's running. curl is a command-line tool used to send data using various network protocols. Here, it's being used to send HTTP requests to the application.

   curl --data '{"user": "user", "query": "How to connect to Kafka in Pathway?"}' http://localhost:8080/
   
   curl --data '{"user": "user", "query": "How to use LLMs in Pathway?"}' http://localhost:8080/

   If you are on windows CMD, then the query would rather look like this

   curl --data "{\"user\": \"user\", \"query\": \"How to use LLMs in Pathway?\"}" http://localhost:8080/

2. Test reactivity by adding a new file: This shows how to test the application's ability to react to changes in data by adding a new file and sending a query.

   cp ./data/documents_extra.jsonl ./data/pathway-docs/

   Or if using docker compose:

   docker compose exec llm-app-examples mv /app/examples/data/documents_extra.jsonl /app/examples/data/pathway-docs/

   Let's query again:

   curl --data '{"user": "user", "query": "How to use LLMs in Pathway?"}' http://localhost:8080/

Step 5: Launch the User Interface:

Go to the examples/ui/ directory (or examples/pipelines/unstructured/ui if you are running the unstructured version.) and execute streamlit run server.py. Then, access the URL displayed in the terminal to engage with the LLM App using a chat interface. Please note: The provided Streamlit-based interface template is intended for internal rapid prototyping only. In production use, you would normally create your own component instead, taking into account security and authentication, multi-tenancy of data teams, integration with existing UI components, etc.

Bonus: Build your own Pathway-powered LLM App

Want to learn more about building your own app? See step-by-step guide Building a llm-app tutorial

Or,

Simply add llm-app to your project's dependencies and copy one of the examples to get started!

Showcases

• Python sales - Find real-time sales with AI-powered Python API using ChatGPT and LLM (Large Language Model) App.

• Dropbox Data Observability - See how to get started with chatting with your Dropbox and having data observability.

Troubleshooting

Please check out our Q&A to get solutions for common installation problems and other issues.

Raise an issue

To provide feedback or report a bug, please raise an issue on our issue tracker.

Contributing

Anyone who wishes to contribute to this project, whether documentation, features, bug fixes, code cleanup, testing, or code reviews, is very much encouraged to do so.

To join, just raise your hand on the Pathway Discord server (#get-help) or the GitHub discussion board.

If you are unfamiliar with how to contribute to GitHub projects, here is a Get Started Guide. A full set of contribution guidelines, along with templates, are in progress.

Coming Soon

• Templates for retrieving context via graph walks.
• Easy setup for model drift monitoring.
• Templates for model A/B testing.
• Real-time OpenAI API observability.

☁️ Hosted Version ☁️

Please see cloud.pathway.com for hosted services. You can quickly set up variants of the unstructured app, which connect live data sources on Google Drive and Sharepoint to your Gen AI app.

Need help?

Interested in building your own Pathway LLM App with your data source, stack, and custom use cases? Connect with us to get help with:

• Connecting your own live data sources to your LLM application (e.g. Google or Microsoft Drive documents, Kafka, databases, API's, ...).
• Explore how you can get your LLM application up and running in popular cloud platforms such as Azure and AWS.
• Developing knowledge graph use cases.
• End-to-end solution implementation.

Reach us at contact@pathway.com or via Pathway's website.

Supported and maintained by