Since the arrival of OpenAI's ChatGPT in late 2022, artificial intelligence (AI) and machine learning (ML) have become ubiquitous topics in the business world. Although the transformative potential of AI has long been recognized by industry insiders and practitioners, ChatGPT has brought it to the forefront of everyone's attention with the force of a prime Shaq going for a slam dunk – with no regard for human life.

Applied Machine Learning is not a rare commodity anymore - there are tons of ML across all industries, ranging from corporate fraud detection, through speech-to-text analyzers, to creating the optimal selfie filter for TikTok. Small to medium-sized businesses are increasingly looking to gain an edge over their competitors by leveraging AI, while in large organizations data science teams are often an integral part of their business intelligence units.The industry is no longer in its infancy - it's $120 billion strong, and it's here to stay. And yet, despite numerous AI unicorns and all the publicity, we believe we are still in the early days.

The industry's quickly learned that the lifecycle of an ML model is not so different from that of traditional software - except being much more complicated. Much like the software industry has developed the need for DevOps engineers, to bridge the gap between software engineering and infrastructure, so too has ML identified the need for MLOps engineers. And they sure have their hands full, as transitioning from running a small, locally built model to a full MLOps pipeline is no small feat.The big fish in the pond (AWS, Azure, GCP) are of course aware of this trend, which is why their ML-targeted offering is growing each year. Various other companies, big and small alike, try to focus on one of the multiple areas of this process to make the life of MLOps engineers easier and allow extracting business value from ML easier, faster, and cheaper.

We believe there is still a massive void in the market for ML tooling and products, which is why we've decided to launch MLNative - the definitive answer for all ML problems that companies face at the production phase of a model's lifecycle. We exist in a vast ecosystem of competing frameworks, libraries, and approaches, yet little to no products that truly solve the issues we've identified based on our experience within the industry.

The cross-cutting nature of MLOps requires effective communication between team members with completely different backgrounds. Due to heavy GPU usage and widely varying deployment requirements, infrastructure management is far from trivial. Finally, as the market takes a downturn, so too are companies looking to save money wherever they can. We're here to remediate all of these problems, and it's only the first step of our journey.

MLnative is going live in June of 2023.

Tomek Grabiński & Łukasz Myśliński