One of the most challenging topics in decentralized finance infrastructure is for sure that most (if not every) smart contract relies on reliable off-chain market data. While Chainlink serves as the oracle on the Ethereum chain, the Pyth-Network seeks to do so on the Solana chain. When we think about trading processes latency and speed is key. The Pyth-Network is able to leverage the speed of the Solana chain (compared to the rather "slow" Ethereum chain) and provide low-latency consensus market data which is imperative towards an infrastructure where market data is not a heavily monetized property of centralized vendors anymore but available to everyone at minimum latency.

Creating an application that is able to consume market data from the Solana chain with the look and feel of a real-time data feed we know from the largest vendors in the industry.

I use the python pythclient for asynchronous requests of the Solana price accounts together with a panel Bokeh server for visualizing the data in a chart:

Please note that non-crypto assets such as AAPL will show a 0 price outside regular trading hours (equities do not trade 24/7 yet ;))

In order to start the panel:

panel serve gridview.py

This part is necessary because of the nested async structure created by the PeriodicDataFrame and the async call of the Solana chain:

import nest_asyncio
nest_asyncio.apply()

The Pyth-Network documentation was of great help in setting this up. Please also see the whitepaper.