Just wrapped up a full RFM customer segmentation project ๐
Took a genuinely messy retail dataset (missing IDs, inconsistent dates, dirty prices) through Excel cleaning โ SQL analysis โ Tableau dashboards.
Biggest surprise: one customer with 88 orders (highest frequecy) didn't even crack the top 100 by spend (frequency โ value).
๐ Check it out on GitHub:
https://t.co/zjBFHOi019
๐Live dashboard on Tableau:
https://t.co/Mi95Ui3pkp
Don't limit yourself based on what you CURRENTLY do.
What you do is not who you are. Some of the best people I've worked with in the Data world were Teachers, Accountants, Warehouse Workers, and Nurses in their past jobs.
I used to be a Recreational Therapist. But I found a new passion and something that I was good at.
So don't limit yourself based on what you're doing right now. If you have a passion for something, go for it!
Oluwafemi Ogunleye is a First-Class graduate of Civil Engineering from Obafemi Awolowo University, where he earned an outstanding CGPA of 4.61/5.00. He is a first-generation university graduate with a strong foundation in structural analysis, construction materials, and sustainable infrastructure.
Oluwafemi's research centres on developing low-carbon, high-performance construction materials to address the environmental challenges of the energy and construction industries. He currently engineers projects across those same sectors, applying his technical expertise to infrastructure development. This is work that keeps his research firmly grounded in the practical demands of the field.
Beyond his technical work, Oluwafemi is committed to service, volunteering with non-profit organisations, mentoring young people, and sharing knowledge within his community.
In the long term, he aspires to pursue advanced research and contribute to the global body of knowledge on sustainable construction, with the goal of becoming a leading academic and researcher advancing greener, more resilient infrastructure.
Seeing all these noise makers who have not created a single job, yet youโre attacking Femi just because you spoke with one or two riders.
Itโs a game of numbers, like he said, and itโs possible for a rider not to earn over 100,000 naira per week.
The bigger problem is that we make it hard for founders building for Nigerians. There is nothing Chowdeck can do about riders sweating profusely when bringing your food. Are they supposed to control the weather or put AC in their bicycles?
Iโve seen Uber Eats and Deliveroo riders deliver food soaking wet just because itโs raining. Should we blame Uber Eats or Deliveroo for that?
If a rider is using a bicycle, thatโs what he can afford to handle his deliveries, and thatโs not a Chowdeck problem.
What Chowdeck has done is provide an avenue for that person to earn a living with the resources available to them. We need to stop looking for reasons to tear down founders who are actually creating opportunities and building for our market.
๐ฒ๐๐ฒ๐ฟ ๐ฎ๐ฝ๐ฝ๐น๐ถ๐ฒ๐ฑ ๐ณ๐ผ๐ฟ ๐ฎ ๐น๐ผ๐ฎ๐ป ๐ฎ๐ป๐ฑ ๐ด๐ผ๐๐๐ฒ๐ป ๐น๐ฒ๐๐ ๐๐ต๐ฎ๐ป ๐๐ผ๐ ๐ฎ๐๐ธ๐ฒ๐ฑ ๐ณ๐ผ๐ฟ, ๐ป๐ผ ๐ฟ๐ฒ๐ฎ๐น ๐ฒ๐ ๐ฝ๐น๐ฎ๐ป๐ฎ๐๐ถ๐ผ๐ป ๐ด๐ถ๐๐ฒ๐ป?
there's a whole system behind that decision, and most people never see it.
we just wrapped the excel series at data with danny cohort 9 with a guided project that builds exactly that system. 20,000 loans, ยฃ548M in portfolio value, real risk modeling, the kind of analysis a bank's risk committee actually sits down and reads.
my students found that borrowers with severe debt burden default at 8 times the rate of low-risk borrowers. that's the headline number. that's the one that gets you a good grade.
but the real skill showed up in what they almost got wrong.
a location table had a column labeled "portfolio value" that actually summed to the portfolio's value at risk. wrong label on a real number. easy to miss, easy to misquote in a boardroom.
and the one that almost shipped without a second look: a "top 10 loan officers" ranking sorted by default rate. looked exactly like a performance ladder. we ran the actual statistics, confidence intervals across 40 officers, and found only 3 were statistically different from the portfolio average. the rest of that sorted list was noise dressed up as a signal.
neither catch was dramatic. each one was a five minute check that would've taken five minutes too long to skip.
that's the actual job. not finding a number. knowing which numbers to distrust before they go in front of a room full of decision makers.
proud of what this cohort built.
#Datafam #Excel