What do cranberry farmers, bugs, deep learning and a mid-western creative agency have in common? Join me to learn how with today’s tools, and a sense of why not, you can jump in and help solve real world needs. In this session, I am going to dive into how we are using deep learning to help cranberry farmers in the mid-west solve real world problems. A part of it is technology, but an equal emphasis is on the business needs that deep learning can help solve. Hopefully I can help show that in today’s world, the obstacles of solving hard technical problems is just the willingness to try something. Key Takeaways: – Data is still king and it helps if you can control collection and labelling of the data. We built an app for that. – You need the right partnerships. Getting SME’s outside your organization to get excited and contribute can really increase the odds of success – You do not need to invent your own models. Sometimes using a model developed for another domain can be adapted for your use case. – There is a tradeoff between accuracy and usability. You don’t need to sacrifice either, but finding the sweet spot can move an idea from an idea to an actual product. – You can take a conventional team of developers and get moving or you can wait and never get started. – It will be very hard to compete against Google and the other gorillas when it comes to general problem domains’ focus on a domain that you have an intimate understanding. Therein lies paths to innovative solutions.