How should AI be utilized and what are the consequences for the public, not just businesses, of its use? The increasing prevalence of ChatGPT has heightened public awareness of Artificial Intelligence and has given rise to a discussion about AI morality.
AI applications have a built-in prejudice, which means we need to be cautious when venturing into this novel AI world since it is not only about how the algorithm is developed and by whom, but also about how the model is constructed.
Notably, AI systems have exhibited various forms of prejudice in public. For instance, Sam Altman, CEO of OpenAI, acknowledged in the last month that ChatGPT has “flaws regarding prejudice”. But if those errors are used in fields such as insurance or pharmaceutical research, the consequences of making bad decisions could be monumental.
MLOps is a combination of Machine Learning and DevOps which is used to deploy and manage ML models in production reliably and effectively. It is a collaboration between Data Scientists, DevOps, and Machine Learning engineers to move an AI algorithm into a functional, operational model. The main goal of this is to automate the model while meeting business and regulatory demands regarding bias and other components of AI. Making processes more efficient has a beneficial effect on the environment.
Seldon, a British startup, is experienced in the specialized domain of development tools that enhance machine learning models. Rivals of Seldon include Arise, Fiddler (which has raised $45.2 million), Dataiku (which has secured $846.8 million in investments), and DataRobot (which has raised $1 billion).
In 2020, Seldon’s platform for machine learning deployment that works with any cloud provider was given a Series A investment totaling Â£7.1 million by AlbionVC and Cambridge Innovation Capital.
A new investor, Bright Pixel (formerly Sonae IM), has led a $20M Series B funding round, with existing investors AlbionVC, Cambridge Innovation Capital, and Amadeus Capital Partners also taking part.
Alex Housley (CEO) and Clive Cox (CTO) of Seldon state that their open source frameworks have seen a fourfold increase in year-over-year growth since their Series A in November 2020. This is noteworthy since the open source network lets them spread their exclusive solutions much more economically and productively.
“Bright Pixel’s Director, Pedro Carreira, stated that Seldon has distinguished itself by providing a special solution that can decrease the difficulty for users to execute and describe ML models in any sector. This leads to increased output for their customers, quicker time-to-value in addition to governance, risk and compliance functions,”
Seldon’s clientele currently consists of PayPal, Johnson & Johnson, Audi, Experian, and other firms.
During an interview, Alex Housley, Seldon’s Founder and CEO, stated that AI is present in every aspect of life and Seldon is in a one of a kind spot. He continued to say that the company is already well-established in its open source distribution, and that they have just confirmed a new approach to MLOps that is focused on data flow and production. In simpler terms, AI models can be enhanced through their algorithms, but the changes are minor. In contrast to other solutions, we have chosen to maximize performance by enhancing the production of data quality. We have had great success with this in our collaboration with Cambridge University.
Based on the survey from Run:ai, it has been found that 88% of companies have experienced a situation in which more than half of the models created never get put into production. The main reasons for this are because projects have been stalled, or there have been multiple teams working on the same task without coordination.
Seldon asserts that its services will enable teams to work collaboratively to reduce deployment time by an average of 84%. This could be of critical importance since more regulations are being established for AI (for instance, the EU AI ACT and the US EEOC). Seldon and other companies in the same field are in a race to assist businesses in complying with these regulations while at the same time enhancing their AI models internally.
The company has worked in conjunction with Neil Lawrence, who is the first DeepMind Professor of Machine Learning at the University of Cambridge.