permutable.ai (formerly known as saffronblue.ai) creates improved AI solutions by combining human and machine intelligence. Enabling humans and machines to work better together, we make AI faster and more accurate.
Permutable’s approach to AI is different because we use human feedback loops within auto-ML and need far less data than traditional AI to operate.
Humans are essential in training and testing data which leads to better AI models. By introducing human intervention to the auto-ML process, models learn more quickly, are more accurate and take less time to develop. Our models only require a fraction of the data normally required which means we can create AI solutions for difficult to automate situations.
Our AI models need around 1% of the data usually required by machine learning models. This means that the benefits of AI can be achieved in situations where data is scarce or hard to gather.
By including human feedback within auto-ML, via a continuous feedback loop, the process becomes smarter, more confident and faster than usual.
One of our first users is Petroineos, one of the world’s largest oil companies, which uses Permutable to automate high-level human tasks within trading. Our software usually delivers a 5x return on investment by way of internal cost savings or boosted productivity.