Superficially, the title of the blog looks like an oxymoron. We immediately map the terms Research & Development with the Innovation. It is so much prevalent in our perception that innovation comes straight out of R & D, ready made, ready to be consumed and everlasting.  But when you carefully look at the process of R & D, it is just an iterative process where you form a hypothesis, conduct experiment and test the theory with the obtained result. If the results are conclusive, then you get your innovation. But it’s most often that you can be wrong. When you fail at the intermediate steps of the experimentations, you go back, re-formulate the hypothesis and test again. It’s called as rinse and repeat. This process demands the sheer amount of energy of the people involved and requires time to do that.

A few years back, while I was learning Machine Learning, NLP, etc., I happened to talk to one of cousin doing his Ph.D. thesis in agricultural sciences. My question was simple, what happens if your experiment fails?  His answer was I will have wait for another year. It takes him a year to conduct one cycle of trials. While back in my previous job, we have developed software system just to cut down the cost of launching new experiments to a couple of days. The framework for starting a new experiment has enabled us to create new experiments at a faster rate, test them and fail fast. It has yielded us a tremendous result. When I look back and reflect on it,  I realize how important it is to shorten the cycle of R & D to bring a positive effect.

Why does the cost and time of R & D is so crucial to businesses?  High Tech business depend mainly on R & D to succeed. After all, they depend on the continuous improvement more than other traditional business. The landscape of these businesses are changing rapidly, and the average age of companies are reducing at a faster rate. Majorly because new innovations are popping up now and then. With all the stakes involved in innovating, one can not just increase the expenditure on R&D too much. The risk of not yielding result should be taken into considerations and mitigate them. Other than the risk of failed experiment cost, the time it takes to test the hypothesis is significant. One of the key factor involved in the success of any product is the go-to-market strategy. How soon can you launch a new product or a feature and get the feedback from the market?  How quickly can you iterate from there? In such case what one can do? Can the businesses reduce the time and cost required conduct an experiment? If so how can they?

Unlike other software development steam, machine learning does concentrate more on research part. It is not unheard of that it takes a couple of months set the baseline for experimentations and few months from there to arrive at the result that is acceptable. If your business wants to invest in such emerging technology or betting on that the next big thing happens will happen through machine learning, how would you minimize the risk? Use the rapid prototyping and development tools that can accelerate the innovation. What would be best if such RAD tools can produce the production-ready codes!

One of the main reason to work on the {{cta(‘512eeaa1-d861-4811-9a17-d8e56c9ce2e6’)}} was to reduce the time it requires to build and ML solutions. The {{cta(‘eaa52814-dc5b-4a31-868e-996d4d3f14ec’)}} platform is powerful that you can create a reasonably complex application in a matter of few hours. The platform helps you run to build the ML app from your browser without having to code and deploy them in our hosted and managed servers. And, use in production straight away. I am reasonably confident that platforms like Datoin will improve and accelerate the innovations across the business.