South Africa

Machine Learning: An Emerging Career Field in South Africa

Machine Learning: An Emerging Career Field in South Africa

Artificial Intelligence as a whole is fast growing  and, in that space, Machine Learning as a career is booming.  

Today, companies collect huge amounts of data, especially about their customers. Machine Learning takes that information, analysis it with the help of computer algorithm to make data-driven recommendations and decisions. The data could be text based but also location, image or voice based. Whenever Google, YouTube, Netflix or Amazon recommend you something, they use Machine Learning. 

Behind this technology are people who can build, repair and maintain these systems. Demand for them is at an all-time high.  Where a data scientist will analyze collected data to identify valuable, actionable insights from a database, a machine learning engineer will design the self-running software that makes use of that data and automates predictive models. 

Since machine learning engineers sit between different disciplines of IT, when trained correctly, they have a foundational knowledge in software engineering principles which is combined with data science in order to produce models that become valuable software. This means that machine learning engineers need to have a slate of skills that span both data science and software engineering.  

This post looks at what essential skills every Machine Learning engineer will need for success in their career field.   

Technical skills needed:  

  • Software engineering skills. Some of the computer science fundamentals that machine learning engineering rely on writing algorithms that can search, sort, and optimize; familiarity with approximate algorithms.  
  • Data science skills. Some of the data science fundamentals that machine learning engineers rely on include familiarity with programming languages such as Python, SQL, and Java.  
  • Advanced machine learning skills. Many machine learning engineers are also trained in deep learning, dynamic programming, neural network architectures.  

Soft Skills needed:  

  • Communication skills 
  • Problem-solving skills 
  • Time management 
  • Teamwork 
  • Thirst for learning 

Now let’s take a look at the kind of tools that your typical machine learning engineer would use. Amongst the programming languages used are Python and SQL.  

South Africa is making efforts to stay at the forefront of developments in Machine Learning, as well as working to solve some of the challenges that we face in this space. 

One of the local drivers of change in the industry is Mr Vukosi Marivate, the Chair of Data Science at the University of Pretoria and co-founder of the Deep Learning Indaba. Marivate has been working on projects to improve tools for and availability of data for local languages. 

Its purpose is to monitor and analyze the use of African languages. The goal is to train AI to convert English to African languages and vice versa more successfully and accurately, as well as using this AI in other ways to make the internet as a whole more accessible to African language speakers.  

In 2013, a local group of industry practitioners and researchers began Data Science Africa, an annual workshop for sharing resources and ideas.  

The shift to making Africa a location and participant in AI conversation is a positive one and will ensure that local content and languages are considered and job opportunities created in this dynamic space.  

Time, scope and cost: the three pillars of Project management.

Time, scope and cost: the three pillars of Project management.

It has been given many names – the Project Management Triangle, Iron Triangle and Project Triangle, we shall be referring to it as the three pillars of Project management. If the number of names they go by doesn’t give you the impression of how important they are then let’s begin to breakdown how these three pillars are the most important concepts in Project management.

When used in combination with effective project management software, the three pillars can give you the ability to drive your projects to success. It also means that the success or failure of your organizations projects will depend on the deadlines, features and the budget set by the stakeholders and upper-management.

As a project manager you have to juggle between all three pillars and try to find the best combination of them for your specific project development process needs.
All three pillars are connected and if you want to change something with one of them the other two pillars would be affected.

Understanding that the three pillars must be kept in mind throughout a projects lifecycle to assist the team to adapt to all changing conditions that project would face in the day-to-day. A team lead by a project manager utilizing this strategy will be able to face numerous obstacles that come their way and their performance will remain consistent.

Now let’s take a closer look at the pillars individually:

Number one:

Time – One of the most important elements that project managers have to consider is keeping track of time a project is taking and will continue to run for. As each task and process that is performed by the team has to be part of the overall schedule.

Wondering what the process of setting this timeframe and schedule here’s an outline:

  • Plan schedule management
  • Sequence all of the different activities
  • Outline each of the activities that will be performed in the project
  • Decide what resources will be needed in the project
  • Estimate the time it will take to complete each activity
  • Develop a full project schedule after analysing the previous steps.
  • After creating the schedule, you’ll need to manage the schedule to make sure that your project remains on track and as a result will succeed

The schedule is the time allocated by the project manager in relation to the requirements set by the stake-holders or upper-management. If a project is unable to keep to its schedule and deadlines then it can be in a lot of trouble.

During the Project planning phase a project manager will outline the time required for various tasks, the time planned is dependent on the market requirements of the project and how fast the project is needed to complete to launch, meet the demands of a customer or to begin a new phase of operations.

Number two:

Cost – Another important element related to the Project management. This is the cost and budget of all tasks related to task/project at hand. What’s most important is cost estimation of the various components.

Some methods of estimating the cost of your project:

  • Using Historic Data: Managers can estimate the budget of the project by measuring different stats from the old and new data that they get from the external market and previous completed projects.
  • Use Bottom-Up Approach: Managers can also use the bottom-up approach to estimate the budget of the project by tracking the lowest to highest budgets spent on previous projects.

Number Three:

Scope – The third pillar is arguably the most important part of the process. This is because all other planning and cost estimation

Towards the end goal set.

You can see the scope as a project manuscript that includes each small detail of the project and how it will run from start to finish. It will also include various risks and threats to the project in terms of completing it within time and budget.

Now having identified and discussed the three pillars we can see that they are vital to the process of a project manager and the importance of controlling them to ensure the success of the project.

Tips To Nail Your Internship

Tips To Nail Your Internship

Written by Elbie Liebenberg for W24

Getting an internship in South Africa’s job market is just as tough as landing a permanent job. Should you be lucky enough to get this opportunity, make the best of it. Here’s how:

Principal of Oxbridge Academy  Elbie Liebenberg says the first rule of being an intern is to take your professionalism to the next level.This means dress neatly and appropriately for the role, always be punctual, be willing to learn, and complete your tasks to the best of your ability, even if those tasks seem boring. 

Apply yourself in every task

Always show that you are interested in learning by asking questions, offering assistance, and giving input where you can. And don’t be afraid to speak up about your ideas.The more you contribute, and the more you complete tasks successfully, the bigger the likelihood that more and more work will be passed your way. By the end of your internship, you will have made yourself a useful part of the team, and employers may decide that you will be a valuable permanent addition.

“Even if that doesn’t not happen, you will be assured of a glowing recommendation, and you’ll have learned important skills that will stand you in good stead when going to interviews and when starting your first real job.” she says. 

Find a mentor 

An internship allows you to build professional relationships with people in your field, which means that even if you don’t land a position at the company where you are interning, your new contacts are likely to alert you to any suitable opportunities that arise in the industry. It’s also a good idea to find someone who has been in the industry or the company for a few years, who can guide you and advise you about your field and about those practical aspects you wouldn’t have learned about during your studies.

Ask for help 

A big mistake some interns make is to think that they are incapable or that they have failed if they are asked to do something and they don’t know how. Nobody expects an intern to be able to do everything that is thrown their way. Everyone understands that an internship is a learning curve, and that it can be quite overwhelming. So if you’re unsure of what it is you are being asked to do, get clarification and assistance straight away.

It is also important to remain open to constructive criticism and not to go on the defence, as this is all part of the experience.

Keep notes 

During your internship, keep notes of all the different tasks you have mastered and the contributions you have made to the team. When your time with a company nears its end, ask whether you may schedule a short meeting with the departmental head or HR.

Then use that opportunity to discuss what went well, and to ask for the leader’s insight and advice going forward.

It’s important at this stage to thank them for the opportunity, and to demonstrate that you are a good fit for the company, that you enjoyed working there, and that you would like to be considered for future opportunities should they arise.

Image sourced here.

Is SA fostering the right skills?

Article written by: Ingé Lamprecht

JOHANNESBURG – According to Moneyweb, almost 20% of decision-makers from South African companies who participated in a new study think their companies will switch to recruiting talent from Africa outside of South Africa or from other continents.

This is one of the findings from a new research report, Domestic to Global Leadership: Positioning South African companies for global competitiveness.

While the trend is consistent with international developments, it suggests that despite high levels of unemployment in South Africa, skills shortages persist in certain areas.

Where local companies also operate on the African continent, 40% of respondents said they would look outside the country for talent. Where they do business globally, the percentage was 30%, says Jenny Tyobeka, founder of JT Executive Coaching and Advisory Services.

The significant complexity surrounding unemployment, education and skills in South Africa was again thrust into the spotlight when students across the country participated in #FeesMustFall protests over the past week. On Friday, president Jacob Zuma announced there would be no hikes in tertiary education fees for 2016.

But tertiary education by itself, whether at reduced costs or not, is highly unlikely to be the silver bullet to South Africa’s unemployment and skills conundrum.

Kay Vittee, CEO of Quest Staffing Solutions, says there is a mismatch between the skills tertiary institutions are providing and the skills the labour market requires.

Yet, a study commissioned by the Centre for Development and Enterprise suggests the issue of graduate unemployment in South Africa is over-exaggerated. The study found that even after the financial crisis, graduate unemployment remained at around 5%.

But, says Vittee, the need for experience cannot be dispelled, and this is why some companies are crossing South African borders to access talent. Firms like Roshcon are importing artisan skills, as these skills are in short supply locally. Some companies also brought civil engineers from outside South Africa to the country in the run-up to the World Cup because of the skills shortage.

Tyobeka says a few years ago Adcorp estimated that there was a supply gap of roughly 216 000 managers, especially in the senior echelons of companies in South Africa. With the available skills, it would not be possible to develop talent in the short-term to cover that gap.

The only way South African companies can overcome this shortage is by recruiting across borders, but also through internal leadership development, she says.

“South African companies are really not doing as well as they should in terms of growing the internal leadership capability. It is a huge problem.”

The exploratory research was conducted between October 2014 and February 2015 through an online questionnaire, involved 79 corporate employees from Top40 and mid-cap listed entities, medium-sized private companies and parastatals and was a collaborative effort between JT Executive Coaching and Advisory Services, Lodestar Marketing Research and Quest Staffing Solutions.

Article sourced: Moneyweb

Output And Wages Linked To Prosperity

Rising labour productivity leads to higher employment, income and profits

Article written by: Luke Jordan, Laurence Wilse-Samson

We are a few months into the season for collective bargaining. Most headlines are and will be dominated by nominal wage rises, which sit alongside gross domestic product (GDP) growth, inflation, the deficit and interest rates as the economic indicators discussed most often.

What is less likely to be discussed is the number that sets the parameters for all of those – growth in labour productivity. That refers to the amount of output, net of inputs, produced by each hour of labour. It measures how quickly a given amount of work is producing more goods and services. GDP growth is simply growth in productivity added to growth in hours of work.

In turn, how quickly output per hour of work grows tells us how quickly hourly wages can rise without stoking inflation. It arises when there’s too much money chasing goods. When productivity is high, and goods are growing quickly, the looser monetary policy can be – and still maintain stability.

Productivity also has a decisive influence on the relationship between wages and competitiveness. If it is stagnant or declining, even below-inflation wage rises may reduce competitiveness. Conversely, if productivity is rising quickly, the effective cost of labour may decline even as wages rise rapidly. This belies the contention that lower wages are “necessary for competitiveness”, that wage rises will “necessarily hurt exports”, or that above-inflation wage increases are “inflationary”. It depends on productivity.

It is also not the case that productivity growth necessarily reduces employment. Greater efficiency will reduce the number of employees needed to produce the previous amount of goods and services.

But the keyword is “previous”. The gains from efficiency will be divided between labour, management and capital, depending on their respective bargaining power. The share to labour increases wages and creates additional demand for goods and services, as does that to management and the owners of capital, if it is reinvested. Greater efficiency leads to greater demand, and employment, wages and profits rise together.

That cycle can break down. The surplus can be captured in part or whole by management and capital. If that happens under tight monetary policy and low fiscal incentives to invest, an economy can end up in a vicious cycle where productivity growth is slow, wage growth even slower (in real terms) and unemployment still rises or stagnates.

That is what is occurring in South Africa today. But that does not mean we should reduce productivity growth. That would close the only route to a high-wage economy. The answer must be to fix the causes of a skewed distribution of productivity gains and stagnant demand, while accelerating productivity growth.

That is what happened in the post-war United States and Europe, and in rapidly growing Asia. It has happened in China, where wages have risen on average 15% a year for the past decade. In most industries in China, nominal wages now exceed those in South Africa. But productivity has risen just as quickly, so the quadrupling of wages in a decade has neither hurt exports nor stoked inflation.

Unfortunately, in South Africa productivity growth has performed very poorly. Long-run data is hard to find – in itself an indicator of low priority – but in the past few years it has grown at a mere 1.3% to 1.5%. This is low compared with countries at similar levels of development.

Productivity, more than GDP growth, is perhaps our central economic problem. It is arithmetically impossible for us to become a rich country if wages do not grow significantly faster than inflation, but they cannot do so unless productivity grows far more quickly. As long as it does not, our largest policy battles will remain lose-lose. Productivity sets the constraints for many of the decisions that we make in the economy. It’s the way to grow the pie.

If this is our central challenge, what can be done about it?

Labour productivity is most usefully broken down into two components: capital deepening, or the amount of capital per unit of labour; and total factor productivity (TFP), which measures how much value a unit of mixed capital and labour produces.

Capital deepening is principally a question of investment. For us, that means greater domestic investment. That is a far from trivial task. In South Africa, it is likely to require changes to the tax structure, more effective industrial policy, and a change in the monetary policy mandate. It also means a stable electricity supply and greater infrastructure investment overall.

But in the long run total factor productivity counts most. It is determined by the technology embodied in equipment and structures, the quality of labour, and the quality of management. Put more simply: How leading edge are the companies in the economy? How skilled are their workers? And how well are they managed?

These are all deep-rooted features of an economy. They are structural variables that require structural reforms. Given the range of variables, these reforms vary widely, not only in their content but also in the way which of them will make a difference and when. The International Monetary Fund (IMF) has recently published a useful review of the evidence, categorising structural reforms by their likely short-, medium- and long-term effect.

At the top of the list are research and development, increased competition and higher skills. All three areas are ripe for reform in South Africa. In 2012-2013, we spent 0.76% of our GDP on research and development. That is not terrible for a middle-income economy, but it can be improved significantly. The developed world and rapidly growing developing economies, such as China, spend between 1.5% and 2% of GDP on it, whereas the United States spends closer to 3% and Japan and Korea 3.5% to 4%.

We can and should try to double our spending. There are already tax breaks for private research and development, which should be maintained, while doubling the government’s contribution from R10-billion a year to R20-billion. Although high-risk research will always lead to some failures, most studies of research and development find double-digit rates of return when evaluated across portfolios.

Second is competition policy. Vigorous competition can help remove bad managers, and reallocate resources to more productive firms. Many of South Africa’s markets still feature large entrenched players. Too little has been done to encourage new entrants.

Third is skills. We must recognise our legacy. Apartheid deliberately deskilled most of our population to protect the income of the white working and middle classes. Today, though “skills” is an often repeated mantra, both public programmes and public pressure is half-hearted.

A chronic deficiency in skills both pushes up the price for them, thus increasing middle-class incomes, and protects the skilled from putting in more effort.

More money is needed, but it alone will not solve the problem. Skills programmes are fiendishly complex. In the US, private sector solutions – private vocational colleges funded by public loans to students – have led to widespread fraud and soaring student debt levels. Nor are successes easy to replicate – many countries have attempted to replicate Germany’s apprentice model but few have succeeded.

But this is not a counsel of despair. Sometimes effective programmes are hiding in broad daylight. In Latin America, for example, industrial skills programmes routinely fail, but agricultural extension programmes have triggered explosive growth in several countries.

In China, the government makes publicly available for every vocational college the entry scores of students, graduation rates and job placement rates. It vets how colleges set curricula in co-operation with local industry, rather than the curricula themselves.

In South Africa, we would need to experiment. We have the skills education training authorities (Setas), skills development levies, and public and private colleges. But little to nothing is being done about the core problems of information and co-ordination.

At the least, we should establish a free, simple-to-use system to obtain information about a college’s performance, perhaps with one-year and three-year placement rates accessible by phone. This would not be simple to set up, but neither would it be impossible – linking to the South African Revenue Services’ tax systems, for example, could automate the calculation of how many students are on employers’ payrolls.

In many cases, unions play a vital role in making skills programmes work. Skills training requires up-to-date information on the skills needs of a sector’s workforce. It also requires as much, or more, on-the-job training as training in colleges.

Individual employers are reluctant to provide this, because the costs are hard to recoup if trainees leave soon afterwards, public programmes easily become detached from shop-floor needs. Unions can help to square this circle by aggregating information, running programmes themselves, or monitoring and reporting “free riding” by employers or employees.

A final source of skills is foreign imports. As the Chinese leader Deng Xiaoping told the World Bank in 1978, “we need your ideas, not your money”. The TFP and skill spillovers from foreign investment are often far more important than the size of any such investment.

In theory, “critical skills” visas exist. In practice, when academics take one to two years to get visas to take up posts at our universities, we are damaging both research and development and skills. When companies cannot bring in the managers they need, we are damaging organisational quality and technical know-how. We need high-end skills, and we must make it as simple as possible for those who have them to come here.

That leads to tackling two shibboleths. First, labour market reforms. In that same review, the IMF stated that these reforms can be negative for TFP, especially in a monetarily constrained environment. Since the South African Reserve Bank has decided that South African monetary policy is at its lower bounds, such reforms will have little effect unless its mandate is changed to also consider unemployment and not just inflation. This would align the Reserve Bank’s mandate with that of the US Federal Reserve.

The other shibboleth is the idea that rising productivity entails otherwise avoidable automation. This is burying your head in the sand. Automation will affect almost every job category and every country. Agri-processing mills in the heart of the Malaysian jungle are automating; Chinese factories are installing hundreds of thousands of robots a year.

Rather than trying to turn back the clock, energy should be devoted to securing the policies and programmes that will harness technological abundance for higher growth and higher equity, rather than ever more inequality and stagnation.

We should set a national goal to double productivity growth. We need a serious and sustained conversation about how to overcome the legacy of apartheid and the failure of our present programmes to develop skills on a large scale. That conversation needs to be specific, detailed and pragmatic. We need to do all of this now, before our ability to trust and work with each other are poisoned by more years of the bad choices that low productivity imposes.

Luke Jordan is a cofounder of GrassRoot. He has worked for McKinsey & Company and the World Bank in China and India. Laurence Wilse-Samson is an economic consultant in New York. He has a PhD in economics from Columbia. The opinions expressed here are their own.

Article sourced from: Mail&Guardian