Everyone knows that Artificial Intelligence (AI) brings a serious competitive advantage. Think Uber vs. local taxis or Amazon vs. local stores. Uber started small and quickly took a huge share out of the local taxi market. Amazon has the same story – it now dominates the market and local stores are not selling as much anymore. Get the picture?
This is a David and Goliath story. Goliath is big and strong. It seems difficult to fight someone who has such a serious advantage.
Could it become smarter? Why not use AI?
Machine Learning development is cumbersome. It is a multi-step-multi-skill process to make predictive models from data. A large company hires a team of business analysts, data engineers, cloud developers, QA/QC, and data scientists. These professionals bring their method and practice with them. Being human, they also create redundancies and are prone to errors. The team eventually learns to work together and can get a model out to deployment every few months.
Like any other software, a model is also a software. It is not “build & forget”. The models decay and have to be tuned. Bugs have to fixed and improvements added over time. Every model created also creates the baggage that nobody has the time to deal with later.
The models also live in the minds of the developers. And of course, it creates job security. What happens if the person leaves? The next person has to start all over again for the model to continue to make sense.
Small companies have limited resources – it is IMPOSSIBLE for them to run this cumbersome process. They want a simpler way. They need results in days, and not next year!
That’s what pushed Braintoy to figure out a solution. We want AI to be accessible. We believe that it shouldn’t be complicated or expensive. The playing field between big and small ought to be levelled. AI should produce results for you and not take the time and money away from your business. We made a production machine learning platform for that.
The question – how do I start?
Try small stuff that can have a real impact. These are problems that you normally react to, but it would be better if you could be proactive. For instance, if you are a store, what if the returns of your products could be reduced? You already have data for product sales and returns. Why not let machine learning make sense of it? Or what if you are a financial brokerage that recommends products, would it not be nice to know what products could work better for your customers? Your customers will like it – and so will you!
AI is not as difficult as you think it is.
Think about it.