Data Scientist for EdGE Networks, Abhay Kumar, talks about his journey with the organisation.
1) With respect to your experiences, how do you think the Data Science team has helped simplify Talent Decisions?
Ans) Finding a relevant candidate for a specific job description is a complex problem, because currently there are millions of resumes present on diverse job portals. Even if multiple resumes possess the required skills that are required, it might not be highly relevant. This can be addressed by understanding the overall skills required for a job, whilst analysing the entire job description thoroughly. By using a new Deep Learning algorithm, like the Transformer-based models that we are currently utilising, we can now study and understand the context of the job description with respect to the sourced resume.
2) Tell us a little about your journey, here at EdGE Networks.
Ans) So far, it has been a truly remarkable experience working with EdGE Networks. I have worked on numerous problem statements like Information Retrieval, Classification and Language Modelling using a Transformer-based model. So, in hindsight the best thing about EdGE Networks is that you get a chance to work with some of the latest technologies in the field of Data Science.
3) What project are you currently working on, here at EdGE Networks?
Ans) As of today, I am contributing most of my time and effort towards an Information Retrieval problem statement. And I’m solving it with the help of Deep Learning and NLP (Natural Language Processing) techniques.
4) With your expertise, what parallels can you draw between our recommendation engine and with that of an E-Commerce platform?
Ans) An E-Commerce recommendation engine is a totally different problem statement. So, it’s best we don’t compare two problem statements as it will seem like we’re comparing chalk to cheese.
5) What, according to you, are the key challenges faced by a Data Science team?
Ans) I think a recurring problem for Data Scientists and Engineers all over is the inability to obtain clean datasets to solve a problem statement. This is because if you don’t have a clean dataset, all your fancy algorithms will yield poor and erroneous results.