A Full Monty for Python: KX open sources PyKX
KX is one of those companies.
It’s one of those companies with 15 offices across North America, Europe and Asia Pacific with a fairly weighty installed base of users – largely in the financial services markets, but in other verticals too – and with a dedicated software application developer & data science engineering following… but you still don’t know KX like you perhaps might.
The company is behind the kdb+ time-series database and real-time analytics engine.
It states its raison d’etre as a dedicated mission to accelerate the speed of data and AI-driven business with real-time technologies in the most demanding data environments.
This is not a data warehouse play – it’s time series remember? – this is what KX calls its Data Timehouse. Outside of its use in investment banks and hedge fund organisations, it is also used in life and health sciences, semiconductor, telecommunications and manufacturing.
“At the heart of our technology is the kdb+ time series database and analytics engine, independently benchmarked as the fastest on the market. It can process and analyze time series and historical data at unmatched speed and scale, empowering developers, data scientists, and data engineers to build high-performance data-driven applications and turbo-charge their favourite analytics tools in the cloud, on-premise, or at the edge,” says KX, in its core company descriptor statement.
A developer full monty
In a move which could see the KX name become increasingly well known – and in a development which the firm itself describes as a ‘Full Monty’ moment (Ed: is this a reference to sausage, egg, beans and black pudding, or Monty Python, or both?) for both developers and data scientists – the company has now announced PyKX as an open-source download, making the power and performance of kdb+ available to the global Python developer and data science communities.
KX reminds us that the Python programming language has revolutionised (and if that’s too strong a term for you, let’s just say invigorated) data science and machine learning with its ease of use, range of libraries and focus on simplicity and readability.
But, asserts KX, there are ‘faster and better’ technologies for bringing data analytics workloads to Python capable of improving the robustness and quality of data science with better ‘time to production’ stats.
KX CEO Ashok Reddy: A fast and simple way to bring the power of kdb+ and q to analytics.
Providing an authentic and familiar Python developer experience, PyKX offers what is said to be ‘deep integration’ with the kdb+ time series database and q, its programming language.
This is described as an integration designed to allow data scientists and developers to accelerate mathematics and analytics-intensive applications for real-time analytics views across all Python workloads.
All Python workloads
All Python workloads? Yes says KX… and that includes workloads shouldering anomaly detection, predictive maintenance, feature engineering, back-testing, quantitative finance etc.
“Since its creation, the Python programming language has transformed data science. Similarly, kdb, has set the standard for real-time analytics on temporal and relational data at both speed and scale. With PyKX, we’re democratising access to our kdb database, q language and Data Timehouse technologies. Existing kdb users now have access to the wider Python developer community while new users have a fast and simple way to bring the power of kdb+ and q to their analytics,” said Ashok Reddy, CEO of KX.
According to Jonny Press, CTO at Data Intellect, PyKX is a game changer for enterprises looking to put kdb+ alongside Python for developer and data science teams.
“Many of our clients use the languages in tandem but the question has always been where the boundary lies, which parts of the workload do you do in kdb+ and when do you shift to Python? With PyKX there is no boundary,” said Press.
To get started and learn how Python interoperates with kdb, a prebuilt Jupyter project hub gives access to a working Jupyter Notebook pre-loaded with KX software and KX training materials, no installation is required.
Once familiar with the concepts from the self-serve training session, users can download PyKX and use with an existing kdb license, a free trial or a kdb Insights Personal Edition to build the next generation of production-ready data science applications.
PyKX will be available as through Anaconda, the popular data science platform used by 35 million Python developers. It will also be available from The Python Package Index (PyPI), a repository of software for the Python programming language.
