Malevich is a platform for building, iterating and deploying AI-driven product backend flows:
- Vast library of pre-built components shortens the gap between the system design and first run
- You are not limited to our components: easily implement and integrate your own codebase
- Deploy on GPU and production with 1 line of code
Malevich is a Python package that help you to build your components and product flows. Dive deeper into our documentation here. Let's walkthrough how you can create an AI-based product with couple lines of code.
To install Malevich, you should have Python 3.10 and pip package manager installed. Run the following command to install our package:
python3 -m pip install malevichMake sure everything is set up:
malevich --helpMalevich works like a package manager allowing you to install components and use them from within the code. Let's install some components for our pipeline:
malevich install spacy openai scrapeBrowse through our library here and find the components that meet your needs.
The real magic happens when components are interconnected into a more complex and useful pipeline. Let's do it by writing a few more lines of code:
# flow.py
import os
import pandas as pd
from malevich import collection, flow
from malevich.openai import prompt_completion
from malevich.scrape import scrape_web
from malevich.spacy import extract_named_entities
prompt = """
You are a professional journalist. You've received
news containing many references to different people.
Your task is to understand the roles of these {entities}
and write a brief summary about them. The brief should
include the following information:
- Who is the person?
- What is their role in the news?
- What are the main events they are involved in?
Only include individuals for whom there is sufficient
information in the news. Otherwise, omit their names
entirely from the brief.
"""
@flow()
def write_brief():
# Scrape some news from the OpenAI blog.
links = collection(
'News Links',
df=pd.DataFrame(
[
'https://openai.com/blog/sam-altman-returns-as-ceo-openai-has-a-new-initial-board',
], columns=['link']
)
)
# The scraper app will retrieve information from websites specified by XPath —
# a query language that allows extracting information from markup documents.
text = scrape_web(
links,
config={
'spider': 'xpath',
'min_length': 100,
'max_results': 25,
'links_are_independent': True,
'max_depth': 1,
'spider_cfg': {
'components': [{
'key': 'news-text',
# Specify XPath query.
'xpath': "//div[@id='content']//text()"
}],
'output_type': 'text',
'include_keys': False
}
})
# Extract names of entities.
entities = extract_named_entities(
text, config={
'output_format': 'list',
'filter_labels': ['PERSON'],
}
)
# Write a brief about the news using OpenAI API
# to generate text based on our prompt and extracted names.
return text, prompt_completion(
entities,
config={
'user_prompt': prompt,
'openai_api_key': os.getenv('OPENAI_API_KEY'),
}
)
if __name__ == '__main__':
from malevich import CoreInterpreter
# Create a task for writing a brief.
pipeline = write_brief()
# Before running the task, interpret it to make
# the platform aware of dependencies and execution flow.
pipeline.interpret(
CoreInterpreter(
core_auth=('example', 'Welcome to Malevich!')
)
)
# Execute the task.
text, brief =<
7E9B
/span> pipeline()
# Save results.
text.get_df().to_csv('text.csv')
brief.get_df().to_csv('brief.csv')Before executing this demo code, make sure you have the Open AI API key in your environment:
export OPENAI_API_KEY=<YOUR KEY>Now, all you need is a 1 command to run the deployemnt:
python flow.pyYour results are stored in text.csv and brief.csv files.
Running this as a script is cool, but what if you want to integrate this flow into your own system with a production deployment? That's done by adding just one line:
# ...
if __name__ == '__main__':
from malevich import CoreInterpreter
# Create a task for writing a brief.
pipeline = write_brief()
# Before deployment, interpret it to make
# the platform aware of dependencies and execution flow.
pipeline.interpret(
CoreInterpreter(
core_auth=('example', 'Welcome to Malevich!')
)
)
# Deploy the flow
print(pipeline.publish().get_url())By running this script, you will get a link to an HTTP endpoint hosted by Malevich Core, which can be integrated into your application, be it a desktop, mobile, web app, or any other program. Enjoy the ride!