If you want to become a data scientist, then there are various skills you need in your arsenal. Having good knowledge of programming, statistics and data visualisation – not to mention the importance of machine learning – are obvious. But sometimes, the most important aspect is finding a company which fits your needs.
“Someone had to say it,” wrote Jonny Brooks-Bartlett, data scientist at Deliveroo in a Medium post in March on the difficulties those in the industry faced. Admitting he was ‘playing devil’s advocate’ somewhat, Brooks-Bartlett noted that often, expectation did not match reality with some companies.
“The data scientist likely came in to write smart machine learning algorithms to drive insight but can’t do this because their first job is to sort out the data infrastructure and/or create analytic report”
Brooks-Bartlett wrote. “In contrast, the company only wanted a chart that they could present in their board meeting each day. The company then gets frustrated because they don’t see value being driven quickly enough and all of this leads to the data scientist being unhappy in their role.”
So how much of this is true, and how much is knowing when to play politics at the right time? AI News caught up with Antonio Fragoso (left), senior data scientist and technical leader at IT and software development firm Globant, about what makes candidates stand out, and on what can be expected.
Hi Antonio. What does your role as a data scientist involve – and what is your day to day routine (if you have one)?
A typical day for a data scientist would involve working in between business, developers and some other experts in fields like user experience, big data and cloud. We help product owners to set up objectives and directions to construct AI/data products for millions of users (they have the expertise on the core business and we provide the ideas to convert raw data into valuable assets through complex transformations and modelling techniques).
We work side to side with developers and technology experts to translate models into the creation and mature of scalable products.
What in your opinion are the key attributes required to make a data scientist? How important is the understanding of AI and machine learning and its associated languages compared with a few years ago?
Besides great imagination, storytelling and of course proficiency in complex applied statistical modelling techniques (machine learning, deep learning) and coding, I think a data scientist nowadays should be deeply involved and hands on working in big data and cloud technologies, which will allow them to be properly involved in the whole data cycle as building highly standard AI products.
What in your opinion are the most exciting initiatives you are working with at Globant – and the most exciting initiatives the company is putting together around artificial intelligence?
I cannot be very specific about the products we’re creating due to confidentiality agreements, but we’re definitely facing exciting times at Globant and more specifically AI and data science.
We’re working on initiatives with clients across many industries, including education, finance and healthcare. We’re in the middle of a great journey, since we’re helping to construct the cognitive and digital revolutions through the creation of products that will emotionally connect our clients to millions of customers and employees, based on the usage and discovery of complex and very powerful techniques, such as neural networks, in text, language and some other cognitive science related applications.
What are you going to be speaking about at the AI & Big Data Expo – and what message are you hoping the audience takes away from your presentation?
I am planning to share a summary of Globant’s high level methodology for building AI projects, as well as some tips and experience while facing challenging requirements. I’ll do this following some examples from a project where we created and matured a chatbot based on some deep learning latest techniques.