overview

Currently Available Modules

🇨🇴 Versión en español de este documento

What follows is a list of all modules currently available to build Krixik pipelines with. As you review the list and begin to ideate pipeline possibilities, keep in mind that (as long as outputs and inputs match) there are no restrictions on how you can put these together. Repetition, even serial repetition, is permitted. Also keep in mind that it's possible for you to incorporate your own modules and models into Krixik.

Given that the list of Krixik modules—and of the models within them—will constantly grow, we suggest that you bookmark this page.

AI Modules

• Image Caption: Generates a textual caption for an input image.

• OCR (Optical Character Recognition): Extracts text from an input image.

• Sentiment Analysis: Performs sentiment analysis on input snippets of text (i.e. is the text positive/negative/neutral?).

• Summarize: Summarizes input text.

• Text Embedder: Converts input text into numerical vectors. These can then be stored in a vector database to enable semantic (vector) search.

• Transcribe: Transcribes an audio file's contents into text.

• Translate: Translates input text into another language.

Database Modules

• Vector database: Creates a vector database with set of input vectors. Enables semantic (vector) search.

• Keyword database: Creates a relational database of keywords drawn from an input text file. Enables keyword search.

Support Function Modules

• Parser: Divides input text files into (potentially overlapping) snippets in a JSON file.

• JSON-to-TXT: Converts an input JSON file to a TXT file.

Adding your own Modules

• Adding your own Modules or Models: On how to incorporate your own models or module ideas into Krixik.