With a centralised group of analytics and data science professionals, Indium Software comprises seasoned data scientists who handle everything from R&D to Technical Feasibility Analysis for all projects. The company also has de-centralised teams that are specific to region, domain, and projects. To understand the data science hiring process in the organisation, we caught up with Pradeepa Ravindran, Vice President, Human Resources at Indium Software.
Ravindran shares that an ideal data science candidate is like two sides of a coin. “One side is where they should possess functional skills such as business acumen, domain expertise and should be able to come up with innovative business use cases. Whereas, the other side of the coin being technical expertise where the candidate should possess tool & technology expertise, hands-on experience, willingness to be flexible and should possess a high learnability index,” she says. Therefore, the hiring process requires them to have a thorough look at both technical skills and business acumen.
Skill Sets For Data Scientist
While details on the skill sets that they look for in a data science candidate, Ravindran shared few points as below that they look for:
- Writing and optimising queries
- Should know ML algorithms like Regression, Random Forest, Xgboost, Naïve Bayes and SVM very well – implementation and maths of it
- Should know Statistics concepts like probability distributions, hypothesis testing, Central Limit Theorem, metrics like p-value, precision, recall etc
- They should have experience with tools such as SAS, Matplotlib, Spark, Jupiter, NLTK, scikit-learn, Tensorflow, Tableau, Power BI and more.
Ravindran further added that on the educational background front, they mostly prefer to hire candidates from B.Tech, M.Tech (IT OR Computer science), M.Sc Statistics and M.Sc Data Science from renowned institutes like (IIT, BITS, etc.). They also prefer to hire talent with Data science specialisation in PhD, i.e. ML, NLP, Deep learning.
However, when it comes to preference between educational background and skills, Ravindran shared that while they prefer both pedigrees, skills are more important for data science roles.
Interview Process
Data scientist recruitment processes dive into understanding candidates’ technical aspects over the video, Fact to Face, or phone call. The interviewer from the technical team would gauge his/her potential in Data science, i.e. coding, Math, Statistics, ML & DL, Critical thinking, problem-solving ability.
Post clearing the hiring manager interview, candidates are given a technical assessment which is monitored with the help of an ATS tool. “They will then go through a CoderPad test. This assessment is designed to take 3 hours to complete the task given in Data Analytics or Machine learning models,” shared Ravindran. This assessment is to understand the candidate’s presentation and the reproducibility of their work.
As a next stage, the company invites the candidates to visit their office premises for the final stage of the interview process. “We also do another evaluation process by our technical panel on problem-solving/whiteboarding. Finally, our HR discussion happens to check the culture and value fitment,” she added.
Some of the questions that assess a candidate’s algorithms and technology expertise in data science are:
- Which algorithm (Decision tree & Random Forest) has high bias and low variance?
- Moment generating function and probability theory?
- What is a sigmoid and binomial function?
- Write a command for creating an empty data frame with 3 columns?
Besides, the candidates are also assessed for their quantitative aptitude, logical thinking and analytical reasoning skills.
Some of the conventional ways of sourcing data science talent at Indium is through LinkedIn, job portals such as Naukri, Monster, Employee referral program and Walk-in interviews. On the other hand, some of the non-traditional ways are Head Hunting through social media, participating in meetups, GitHub sourcing and connecting with learning institutes to hire potential talents.
Some of the current positions that they are hiring are Lead Data scientist, Junior Data Analyst, Data Analyst and Analytics Data Engineer. Interested candidates can apply on LinkedIn or the company’s career’s page.
Picking The Right Candidate
Ravindran shares that they face several challenges while recruiting for data science positions. “More often, the data science candidate’s resume looks vibrant with a gamut of technical skill sets mentioned in the resume, yet, they flounder on basic concepts of Statistics and Math which is essential. This makes it very difficult to nurture and groom the candidate to take on additional responsibilities when they are conceptually weak,” she added.
Suggesting some of the ways that companies can avoid hiring mistakes are to make it a point to assess candidates on their statistics and math concepts. “This assessment ensures that they can contribute and scale up within the organisation quickly. We also assess them critically on their business acumen and even assess their interest in cross-domain expertise,” she said.
Growth Opportunities For Data Scientists At Indium Software
A data scientist is expected to perform various tasks from designing data modelling processes to creating algorithms and predictive models. They are also expected to create custom data models and algorithms while solving complex problems in statistical analysis, machine learning, deep learning, NLP, CNN and more.
Apart from exposure to a wide range of challenging tasks, Indium provides opportunities for data scientists to grow internally. As Ravindran shares, there are no boundaries to grow at Indium. “With our diverse range of clients, career development opportunities are spread very wide. Even by starting as a Junior Data Scientist, he/she is trained to build strong technical knowledge along with a wholesome understanding of the business acumen,” she said.
“As our Data Scientists grow, they are equipped to take more responsibilities and be accountable for their assignments. We give more importance on the skill and competency a person possesses, and so we have people reach the role of a Principal Data Architect within a period of 10 years,” she further added.
In a concluding note for an analytics professional who wishes to carve out a career in the analytics industry, Ravindar advises, “In order to have a fruitful career in data science, my advice to budding data scientists would be to always have a holistic understanding of your trade – be strong technically and functionally.”
“A one-dimensional way of thinking and working is not enough anymore, be multi-dimensional. The ability to support different functions is an absolute must – it can be from pre-sales to customer/account management. Apart from this, push yourself to be a continuous learner, be articulate and always be a team player. With these qualities, the journey from being a Junior Data Scientist to becoming a Principal Data Scientist can be achieved in a short span of time,” she said.