Applications in Spotlight

Paternova helps some applications more than the others.

Statistical Analysis and Stock Price Tracking

Step-by-step implementation of simple statistical analysis on stock prices in Paternova.

This project is taken from example 3 in Introduction to Paternova.

  1. Using plans provided in Paternova, import daily stock prices from a third-party API.

  2. Calculate SMA of the stock prices with a window size of 5 days.

  3. Compare the results in an x-y chart.

Matrix Operations and Conway’s Game of Life

This project is taken from example 5 in Introduction to Paternova.

  1. A matrix is initialized as the first generation of cellular automata in Conway’s game of life.

  2. A colormap is applied to the matrix view to easier visualize alive cells as black and dead cells as white.

  3. Following the rules of he cellular automata, the alive cells that will remain alive are calculated.

  4. Following the rules of he cellular automata, the dead cells that will come to life are calculated.

  5. Adding these two matrices will create the next generation of cellular automata.

  6. If a loop with a conditional plan is implemented, this cycle can repeat indefinitely by refreshing the cells involved. Paternova has other ways to tweak the refreshing process, for example by refreshing periodically.

Step-by-step implementation of Conway’s game of life in Paternova using matrix operations.

Image and Text Operations with LLM’s

This project is taken from example 6 in Introduction to Paternova.

  1. Text and images are natively supported in Paternova. Let’s first create a custom text as a “prompt”.

  2. Pass this prompt to Gemini LLM through Gemini text generation API implemented by a plan in Paternova. Chat output text is available after a refresh.

  3. Similarly, pass the prompt to OpenAI LLM through OpenAI text generation API implemented by a plan in Paternova. Chat output text is available after a refresh.

  4. Pass the prompt to OpenAI LLM through OpenAI image generation API implemented by a plan in Paternova. Image generation output is available after a refresh.

  5. Let’s send the image back to LM’s for interpretation. First, send the image from last step with along with another prompt to Gemni LLM though Gemini vision API implemented by a plan in Paternova. The text description output is available after a refresh.

  6. Smlalry, repeat the last step by sending the image and the prompt in the last step to to OpenAI LLM though OpenAI vision API implemented by a plan in Paternova. The text description output is available after a refresh.

Step-by-step implementation of text and image operations in Paternova and accessing multiple LLM API’s.

AI-Powered Security System

Designing with Paternova can be broken down into a number of steps. Let’s follow them to implement a security camera system enhanced by AI:

  1. Design Plans: The objective is to design a simple security system that retrieves frames from cameras, identifies humans and birds in them, and records them separately.

  2. Identify Plans: Camera frame retrieval is implemented as a Get operation in Paternova. Frames may need to be resized or cleared before sending it to OpenAI API. Finally, the mages are recorded using Post operations.

  3. Implement plans: In this step, plans are translated into Paternova. This often requires introducing intermediate plans.

  4. Expand plans: The workflow developed in the last step is expanded here. The possible expansions for this example include capturing from several cameras, saving frames to video, recording periodically with refresh timer, posting to email server, etc.

AI-Powered security system and general design flow in Paternova .