Africa is on the rise, with technology such as mobile money helping to fuel development.

Kenya, in particular, has carved out a niche for itself as a technology hub on the continent, with many young innovators creating software and apps.

However, there has been concern among experts in the country of a lack of relevant skills required for the current technology job market.

A study carried out in 2011 by the then Kenya ICT Board revealed that there is a software skills gap in Kenya, despite a high demand for software development services.

In order to mitigate this skills gap, the Kenya ICT Board (now Kenya ICT Authority) has partnered with Carnegie Mellon University to create a software certification examination dubbed ‘Chipuka’.

The certification is set to ensure that developers who take the test are offered IBM-backed credentials to help employers easily identify them as having the necessary international skills.

The exam is set to be officially rolled out this October, and ITWeb Africa caught up with Philip Miller, project scientist at Carnegie Mellon University (CMU) responsible for the development of the certification

ELLY OKUTOYI: What motivated you to initiate the developer certification project?

PHILIP MILLER: I joined the computer science faculty at Carnegie Mellon in 1979. My responsibility was to deliver freshman programming classes. At CMU this meant I taught, or was responsible for teaching, about 1,000 of the 1,200 freshmen who entered the university each year. Students in fine arts, music and theater were not required to take my programming courses. Everyone else was. Programming was red hot in those days and perhaps nowhere was it hotter than at Carnegie Mellon. I soon discovered that cheating in computer science courses was rampant. It seems there were three ways to pass a programming class:
1) Know how to program,
2) Know somebody who knows how to program or
3) Cheat.
In the spring term before I was hired roughly 80 percent of the class submitted an identical solution to a major programming assignment. It was an excellent solution, authored by the course teaching assistant and left it in a publicly readable directory for a few hours. Students cheated. I made it my mission to see to it that a passing grade in programming at Carnegie Mellon reflected the ability to program.

I decided that the best way to ensure that students could program was to have them program while proctored. I drew on my own experience as a music student. Recitals and juries before the faculty left no doubt about who could and who could not perform. I made the final exam, which I called the mastery exam, count for most of the course grade. It was a five hour opportunity for each student to program one of a handful of programs. A passing grade required reading a file into a basic data structure and writing it out. Higher grades required students to implement various algorithms and add additional data structures. Over the years I got more creative, building a base application that students were required to modify. This enabled me to ask both simpler and harder questions, setting a lower bar for passing grades and a higher one for an A.

I'd like to report immediate success and broad acceptance of a level playing field and standards that were both unambiguous and high. That would be a lie. Overall the student body hated the exam and even more than the exam, they hated me. As the most junior of faculty I was called before the Board of Trustees to explain why I was being so hard on the students. I don't recall a lot about the meeting except that at the end the lead trustee demanded to know why CMU didn't test calculus and physics with a mastery examination. I figured I wasn't getting fired that day. In my second or third year at CMU a Computer Science departmental review was conducted. J.C.R. Licklider (famous professor from MIT who conceived of the Internet in 1958) chaired the review committee. I was given 30 minutes to talk about the mastery. No one interrupted. No one asked a question.

My passion at CMU was to share some of the brilliance and excitement of our computer science department. I attended one of the really bad high schools in America and wanted to make it easier for students at the bottom of the pyramid to experience high quality education. This led me to create the College Board's Advanced Placement Computer Science offering. Later, in 1996, I taught the first Internet courses at CMU. I quit my job, talked the Associate Dean into quitting his job, and we started iCarnegie. This company took the software development distillation of CMU's computer science curriculum global. We licensed partnering institutions at roughly 5% the cost of getting the same class at CMU. Bitterly I learned that about 70% of the institutions that licensed our courses had no interest in educational outcomes. They used iCarnegie and its relationship Carnegie Mellon as a marketing tool. Some clients - Tec de Monterrey, the Adelaide Institute of TAFE, City University of Hong Kong, Laureate’s universities and others did a brilliant job with the curriculum. Their students thrived. However most institutions cynically took our name and gave their students little or nothing.

I realised that licensing superb curriculum, careful attention to teacher preparation, and first class mentoring are not enough to cause disinterested owners and dodgy educators to deliver quality education. I was pushing on a rope. I decided that the best way to ensure quality education in software development was for the software industry to demand it. I knew that by authentically examining the programming skills of job seekers I could help industry identify those who were ready to be productive. Once again I quit my job to go on to something more important, building the Software Developer Certification. The Government of Kenya, the World Bank, Carnegie Mellon and most especially the Dean of CMU's School of Computer Science, Randy Bryant, gave me the opportunity realise the vision that was over 30 years in the making. At every turn the software developer exam (SDC) project has gone better than I expected and I expected a lot. I found more recognition of the need among the world's great software companies than I expected. I found better technical solutions for exam delivery, grading and statistical analysis than ever before. Roger Dannenberg, Chad Dougherty and Robert Seacord are responsible for that. We got one super great idea from IBM's Bonnie John - build the exam on open source base code. I found a surprising number of former students eager to contribute to the success of the SDC. I found a serious partner in the Kenyan government and great colleagues in Grace Kihumba and Linet Kwamboka who came to CMU from Kenya to work on the project.

At this moment we are ready to test the world, really. We have testing tools, an exam and processes for creating additional exams that make testing the world the obvious next step, not hyperbole. The payoff can be immense. Companies will be able to identify qualified job seekers at very low cost. Countries will be able to signal the existence of a well trained workforce. Educational institutions will be able to focus on real needs and stick to them until their students are properly trained. A capable job-seeker, regardless of institutional ties, will be readily identified and therefore have real prospects for a career. Students will have an honest and accurate appraisal of their skills.

Jobs in software development are transformational for families transitioning up, into the middle class. They are transformational for the first kid in the family who didn't - cut sugar cane, pick coffee, mine coal, plumb houses, work in the auto body shop, have a job in the steel mill, harvest broccoli and so forth. Our contribution to upward mobility is the software developer certification and the family of exams that we are creating. Testing software people is what I know best. My team is superb. We are open for business.

ELLY OKUTOYI: What led you to specifically choose Kenya?

PHILIP MILLER: I didn't. Kenya chose me. I told the World Bank about my ideas. They liked the ideas but the Bank does not fund university research. They make loans to countries. I had to give my ideas to the Bank, hoped that a country would find the ideas worth pursuing, responded to a competitive Request for Proposals and in an above board procurement process had the Carnegie Mellon proposal judged to be the best. Kenya picked Carnegie Mellon, not the other way round. At Carnegie Mellon I was simply the guy who knew how to create the exams.

ELLY OKUTOYI: What are the current updates on the progress of the project?

PHILIP MILLER: We have achieved all technical objectives of the project. We have built exams that crisply distinguish levels of programming competence, providing lots of detailed information as a by-product. Software companies are beginning to use the exam to find the developers who are best prepared to work. Universities are beginning to use the exam to build competence into their students. This latter point is really important. If the SDE is used only as a summative indicator of skills it will be a help to industry. When it is used as formative evaluation throughout the curriculum it is accelerates learning. Universities, some really big ones, "get" this and are putting the exams to their best use. This will be transformational as I describe above.

ELLY OKUTOYI: Apart from providing a universally accepted certification, what other impacts and goals does Chipuka intend to have?

PHILIP MILLER: I want to change the world, really. Poverty is being unable to make meaningful choices about your own life. I want people to be able to choose to enter software development as a career. All too often this has been a false choice. Over the years I have seen young people and their families making huge personal sacrifices to go to school in hopes that the child would one day enter the modern workforce, only to be bitterly disappointed due to poor education. I want to fix this. When I am done I want bad university education to become a rumor, not a fact. I want people to be able to choose to improve their own lives through education, hard work and economic opportunity.

My earlier attempts were like pushing on a rope. But here in Kenya, if IBM and Accenture and Oracle and the local software companies use the SDE then training will focus on the skills that people need to get software jobs, and to succeed in those jobs. The SDE aligns interests, making it sensible for students, educators and enterprises to all do the right things. What's not to like about that?

ELLY OKUTOYI: Will the examinations be rolled out to other countries after the pilot in Kenya come October?

PHILIP MILLER: Yes, that is a key part of the strategy. We want the global companies to invest in Kenya so that there are plenty of jobs for Kenyans. The way to attract investment is to unambiguously indicate the presence of a well trained workforce. The most convincing way to do that is to have a few thousand Kenyans each year scoring well on an exam that TCS, Infosys, Cognisant,, Accenture, IBM, Siemens etc ... are already using in their own recruiting. An international rollout is key to Kenya's success.