5 tips for getting started with Machine Learning models

By: Ancoris says
Tag(s): Data
Published: Jun 19, 2019
5 tips for getting started with Machine Learning models

Machine learning (ML) is no longer just for a few dedicated ML experts and the companies that can afford to hire them. New tools like Google AutoML and Google Dialog are letting millions of developers and business analysts tap into the power of ML to transform business processes.4 ways to use machine learning in business

If you’re looking to join them, here are 5 tips for getting started with ML models:

  1. Find an ML champion in the business. This is good advice for any project, of course. But with a relatively new technology like ML, you don’t want to be seen as pushing technology for technology’s sake. Spend some time understanding the business mission and priorities, so you can find a real pain point that ML can solve. That will help you recruit a high-level sponsor who’ll give you enthusiastic support.
  2. Build prototypes quickly, in days not months. Use an agile development method — such as Scrum or Lean Software Development — so you can check in often with stakeholders and make sure you’re delivering what they need.
  3. Demo often. Take every opportunity, from staff meetings to stand-ups in the hallway, around the water cooler and all kinds of company events — or even chance encounters with senior managers in the lift. If possible, create a prototype that runs on your smartphone and can be demonstrated in less than a minute, so there’s really no barrier to offering a demo.
  4. Create prototypes that stakeholders can play with, even if you’re not around. Tools like Google Dialog, which lets developers quickly create ML-powered chatbots, can provide a free phone number that stakeholders can use to independently access your prototypes.
  5. Help figure out the production requirements. Your colleagues in operations may be wary of taking responsibility for a production system based on new technologies. Stick around to help them understand the compute, storage and networking requirements — and associated costs — and then help them deliver the production system. Of course, with cloud-based tools like AutoML, it can take just a few minutes to turn a final prototype into a live production system.

With this kind of approach and access to these new tools:

  • finance and insurance companies are able to detect fraud more efficiently
  • retailers can predict the best price point for their products
  • game developers can target ad spend more effectively to attract players who will become paying users
  • media and entertainment companies can increase customer engagement, brand recall and subscriptions.

If you’d like to find out more about how your business can benefit from the latest generation of machine learning tools, come and talk to the data analytics experts in our GCP team.

< Back to resources

Think big. Start now.

We don’t believe in Innovation, we live it. Innovation combined with pragmatism is what runs through our veins. We ask ourselves the same question over and over again: Does it deliver value? And how quickly? Your big ambitions can start now.