My data science career exists due to the appropriately timed convergence of a problem that needed to be solved, my natural inclinations and an environment conducive to learning and experimenting. I did not seek it but I embrace it.Short back story: I was leading a team of IT professionals in a connected car R&D team working with data scientists, who were essentially the end users of the tools and labeled data we were helping generate. The lack of capacity in the data science team meant they could use additional help. My quantitative leanings and curiosity meant that I got a chance to learn and practice data science as a means to solve real problems.The first problem ended up being a good fit for a supervised machine learning solution. Given that Allstate had a well established data science department and infrastructure, I had access to all of the resources from within Allstate in addition to MOOC. However, the differentiator was not the content but rather the guidance I could get from other experienced data scientists solving a variety of problems for Allstate. Among these, brainstorms to formulate approaches and large peer reviews at key stages kept my learning as well as the value of the solution on track. There were several debates around how and why to partition data, underfitting vs. overfitting models, which models to try, how to tune hyper-parameters etc. that simply cannot be replicated by a MOOC only study.The value I was able to return, in addition to the actual work product produced, was my IT and consulting background that helped the team develop along the dimensions of presenting results, building consensus and understanding that we develop solutions and not models. Over the years, I got the opportunity to develop more depth in the discipline by solving a few other problems before focusing a majority of my energy to building a data science team.