Gaurav Chhabra is an experienced IT Recruitment Specialist for Data#3, specialising in the Australian Federal Government sector.

He has an intimate knowledge of the Data Science market in the Australian Public Service, and is highly regarded for his ability to successfully match candidates to specific technical roles.

Gaurav was kind enough to share his insights, and to offer some valuable advice to Data Scientists seeking to enter and grow their career in the APS.

You’re an experienced and successful recruiter in the Canberra Data Science and IT market, with valuable insights as to what Employers seek and candidates can offer. What are some common trends you’re seeing within the Data Science space i.e. growth in the market, demand for specific skills etc.?

Thank you for the kind words, Dr Antic. Although, the macro environment of the employment market in Canberra has many layers of governance to it, particularly from a security clearance point of view, we have seen a modest rise in Department’s wanting to feed organic skills growth.

We are therefore witnessing a marginal number of non-security cleared citizens with good R, Python and C++ skills being considered for security clearance by the Commonwealth, where they have been deemed suitable for a particular role.

Common trends: In addition to a strong foundational base in C/C++ and/or Python and R, candidates who have experience with concurrent and distributed systems are in high demand. Furthermore, candidates who have a background in probability, statistics and data modelling are looked upon favourably as having these skill sets eradicates the need for various other resources who would need to work in conjunction with a Data Scientist/Machine Learning Specialist.

The next big area for growth would be for candidates who have experience in speech recognition or audio processing, and analytics and data processing technologies and architectures.

What can candidates do to better position themselves to get into Data Science? Sometimes with lack of experience, it’s hard for them to get a foot in the door, is there anything specific they can do to increase their chances?

My advice to candidates looking at getting into Data Sciences would be to build a strong Statistical/Actuarial/Economics foundation and then lead into IT. Whilst this is a lot to ask for (seeing that these skills are already in high demand), there is no better way to gain traction in the market than by combining such a foundation with programming languages such as C, C++, Python and/or R.

Another piece of advice here would be, don’t just focus on degrees/certificates; whilst these are essential to having a long, sustainable career, nothing beats hands on practical experience. Volunteer work with charities and other NFP organisations is a great way to give back to the community, whilst gaining hands-on experience at the enterprise level.

Finally, it is not worth artificially inflating rate expectations by giving in to the hype. Yes, there is a significant level of demand in Canberra, however let’s not forget that these are taxpayer funds being utilised and every department does the right thing by taxpayers by conducting due diligence on candidate rate expectations. Rates within reason and in line market rates are justifiable, however if expectations are 20-25% (for example) higher than someone who comes from a very strong DWH/ETL background, establishing (or growing) a Data Science career in Canberra will be difficult.

How realistic are employers expectations of what Data Scientists can offer their organisation? Do they often have clear goals or do they usually need guidance to help set strategic and realistic goals?

As is the case with any market/industry going through its infancy phase, expectations can be somewhat loosely defined. However, most decision makers in the market work with in-house Data Scientists to redefine expectations as their respective Data Science practices mature. As the market stabilises from a skills gap perspective, and Data Science grows, I can see positions like Chief Data Scientist and/or Chief Data Officer taking the strategic definition part from respective business areas. Until then, we will witness many goal posts shifting within the Data Science market.

What do you think the split is between entry level Data Science/analytics roles and senior Data Science roles in the Canberra market?

Presently, I’d say there is a 9:1 (Senior Data Scientist:Entry level) ratio, from a Contracting market perspective in the ACT. The reason behind this is that Canberra has traditionally employed a large number of Data Analysts (SQL et al. skill sets) and we are witnessing these skill sets evolving to the needs of the future.

What are your expectations of the growth of the Data Science and analytics market in Canberra over the next year or so? How serious are Government departments and agencies taking adoption of data driven decision making?

Most discussions had with C level officers and Seniors, such as yourself in the Data Science community, lead me to believe that we are approaching a point where bottom line benefits are becoming realised quick and thick. The Australian Public Service is far more agile than traditional connotations of change within the public sector. In the coming year or so, the outlook is looking strong for further savings on taxpayer dollars by leveraging Automation, Machine Learning and growing Data Science practices.

What are some tips you’d like to share with employers looking to build their Data Scientist capabilities?

“Let Data Scientists drive Data related strategic visions and roadmaps!” I find that there is a disconnect with strategic decisions being made at the business pain point level as opposed to an enterprise wide Data Science Strategy. This may work to fix spot fires, however, greater empowerment means a lower attrition rate and therefore leading to a sustainable Data Science practice.

As is the case with any industry going through a growth phase, skill shortages are bound to be a regular feature. This is particularly acute for the Data Science industry, as there is a historical skill shortage in economics/statistics/mathematics/actuarial sciences, therefore complicating the quest for the perfect Data Scientist. This leads to an increased level of intra-employer competition when it comes to attracting and retaining Data Science related skills.

It is therefore advisable to employers looking at building a healthy and sustainable Data Science practice, that employee retention is kept at the heart of it.

What are some tips you’d like to share with aspiring, junior and senior statisticians Data Scientists looking for their next role, or wanting to break into the Canberra market?

My Advice, broken down for various levels, is as follows:

Aspiring/Junior: No better place than the APS to start your career. apsjobs.gov.au is a vital place for graduates to register, so they are on various Departmental Graduate programmes. Not only will they gain valuable hands-on experience, in some instances, the respective Department may put them through Security Clearance. Some great pathways and courses are in this guide put together by DPMC: https://www.pmc.gov.au/sites/default/files/publications/data-skills-capability.pdf.

Intermediate/Senior: The Commonwealth offers a wide range of programmes to stimulate Data Skills in the industry, and the Data Fellowship programme is an excellent pathway/upskilling platform, further information here: https://www.pmc.gov.au/sites/default/files/publications/data-skills-capability.pdf.

Furthermore, the following excerpt from the above publication from DPMC is highly useful in order to classify the different types of data skill sets. Upon classification, it is best to stay up to speed by working on various technology stacks, the more exposure you gain today, the higher your employability will be tomorrow. For example, Data Infrastructure Engineers should gain as much experience as they can on:

• Distributed systems concepts (CAP, “Fallacies of Distributed Computing,” etc.)
• Unstructured storage (distributed filesystems, blob storage)
• Structured / indexed storage (column stores, faceted search)
• Distributed processing models (map/reduce, RPC services)
• Distributed schedulers and resource allocators (e.g., Apache Mesos and YARN)

Disclaimer: The opinions expressed by Gaurav represent his own personal views and not those of his employer.