Here’s a break down of what we’ve been up to since we went live a couple of weeks ago:
We have realized that there is a huge gap in this space in the ecosystem. There needs to be a pool of people who can work with clusters, crunching data, setting up platforms and so on.
We have began our first SRE (Site Reliability Engineer) training program where we will be training various individuals on network design, platform design, fundamentals of availability, disaster recovery and business continuity, troubleshooting and critical thinking. These engineers will be fed back into the market. The program will last six months and is mainly focused on ICT Infrastructure. The first trainee is Moses Temu
, a student at USIU.
We have a new technical lead at the iHub. We have made great strides as far as the platform is concerned. We have a third node live, increasing our compute power by a further 11 teraflops. The platform is now running on Debian stable. We are now using SLURM as our resource manager of choice and Ganglia for our monitoring.
c)How does the Kenyan Economy Work?
A macroeconomic model
is a tool designed to describe the operation of an economy in a region. This would be a large-scale macro econometric model where we would depend more on past correlations instead of theoretical relations. The idea is to find out what exactly causes our economy to grow/stagnate.
For example, impact of rainfall on electricity output and the net result of that. Correlation between higher transport costs (due to the fact that most of our goods are transported using road and not rail) and reduced exports. What causes the Kenyan shilling to actually fluctuate. By broadening the scope of our datasets, we hope to have a more accurate picture of how the economy in Kenya functions.
We have the infrastructure to do this. iHub Research and iHub Consulting are heavily involved in the planning out and eventual deployment of this. We are working to include a local economics think tank along with other private sector players who would add a tremendous amount of value to the project. If interested in participating or helping out with this project, write to [email protected]
This is the second project that we are kicking off together with iHub Consulting, the Praxis Group (http://www.praxis-group.com) and Brain Trust Strategies. This blog will focus purely on analytics.
We have a few analysts and economists who are looking to find correlations and patterns within large sets of data. The findings will all be published on the blog, to create general interest in analytics as a practice and create use cases for the cluster.
A good example of this, is if we could, for example get data from supermarkets and their customers who purchase using credit/debit cards (you are sure about who is shopping at that point - more accurate than the loyalty card, which can be interchanged with family members) you can get a good idea what people love to buy, when, along with their demographic information. This will put the supermarket in a good position to adapt to who their actual clientele are (for example, you may find out that the supermarket is in an area filled with university students so you can increase stocks of products that will be of value to university students and reduce those that wouldn’t be in high demand). Such use cases of data (open and otherwise) are what will be published on the blog.
There will be a technical deep dive into the platform at the next East Africa Network Operators Group meeting.
We are currently taking requests for projects and applications from the community. If you need your application to run on the cluster or have suggestions for new projects we should adopt, show up for the next EANOG meeting or send an email to [email protected]
, someone will get in touch with you.