Where have I been?
The simple answer...no. No I didn't quit.
But then, where have I been?
Existing mostly. I'm still going strong with my Calculus and Data Science course. I've even started learning French. Then I realized I'm just grinding my wheels in the dirt with no real progress. I actually came to that realization yesterday. After months and months of studying and thinking that I was upskilling, I finally decided to make use of an employment preparation tool offered by my university. It lets you test yourself with interviews and assessment practices. So, I took the first assessment,
and got a beautiful,
red-colored,
D. Fucking D.
I was in the 14th percentile. I have never felt like more of an idiot.
But then, there's something else I've been learning about this past month. It's called, the growth vs fixed mindset. All my life I've had a fixed mindset coz I never really had to try hard to pass. I have never gotten anything below a B throughout my school years. I've always topped my class. Now I'm in the bottom 14th percentile on a global scale. And I finally feel not so smart.
But I can't just quit. Not because I have this crazy, practically delusional willpower to make something of myself. It's simply because, I can't quit. The world I was born into doesn't allow that. I can't depend on my parents forever, I'd feel pathetic doing that. I can't depend on government handouts, my country is too pathetic to have anything that nice. I can't live the rest of my life with no income source, yayyy capitalism.
So it's either I do something about my shitty brain or die. And while not existing sounds very enticing sometimes, I'm not ready for that yet. I'm still a little curious about what the future holds. Not curious enough to live past 23 if my life doesn't turn around within the next few months. And I can't sit down and wait for my life to turn around, I have to actually do something about it. So, here I am, being mediocre but hoping that mediocre is at least good enough.
MobileMuuzaji
The main reason I decided to post today actually is to share a project I have built (mostly vibe-coded because I don't know the first thing about Kotlin) for my mum's business. There's a high likelihood I'll end up working there so I'd better start taking an interest in what happens, and figure out how I can make my Computer Science degree useful on a farm and small shop in the middle of nowhere in Kenya. Also, I love my mum.
I want to write a post about it because I was told this is a good way to get discovered by potential employers. Why did I start by writing a sob story that will most likely put off any potential employers? I don't know. I'm not that smart. I think we've established that already, haven't we?
This post aims to outline the development process and main features of MobileMuuzaji. This is a mobile POS (Point of Sale, not the other one you weirdo) terminal specifically designed for small business owners in Kenya. It allows them to easily track and manage inventory as well as employee details. It also offers data visualization tools so that they can model the benefits and drawbacks of their different inventory items (this is mostly here just to show off my understanding of data visualization :) ).
What Does MobileMuuzaji Actually Do?
Before diving into the analysis, here's a quick rundown of what the app offers at its current stage of development.
The core idea is simple: give market vendors and small shop owners a tool to track their stock and record sales without needing a constant internet connection.
Current Features
Authentication & Onboarding
- A guided onboarding experience for first-time users with contextual tooltips that walk users through the app's key actions
- Email and password signup and login with server-side validation
- Persistent login sessions so users don't have to log in every time they open the app
Organization Management
- Users can create and manage organizations (think of these as individual businesses or shops)
- Organization admins can add employees by email and remove them from the organization without deleting their accounts
- Both admin and employee roles are supported, with role-based access — for example, only admins can manage employees
Inventory Management
- Add new products with a name, quantity, unit (supports both weight and volume units — kg, g, lb, oz, litres, ml, etc.) and cost per unit
- Edit existing inventory items
- Search inventory by item name in real time
- Offline support — inventory changes made without internet are saved locally and automatically synced when connectivity is restored
Sales Recording
- Record sales directly from the inventory list using a "Sell" button on each item
- Two input modes: enter how many units were sold, or enter the leftover quantity and let the app calculate the amount sold
- Built-in stock validation — the app prevents recording a sale that would exceed available stock
- Offline support — sales recorded offline are queued and synced to the backend automatically
Sales Analysis
- View sales history per organization
- Filter sales by date (today, this week, this month, all time, or a custom date range)
- Sort by date, alphabetical order, or earnings
- Toggle between individual sales view and a grouped view that aggregates sales by product and shows total earnings per item
Offline-First Architecture
- The app uses a local SQLite database (via Android Room) as a cache
- All data is pre-loaded on login so the app remains fully functional without internet
- Background sync via WorkManager pushes local changes to the PostgreSQL backend whenever connectivity is available
- Offline-created items use temporary IDs that are replaced with server-assigned IDs on sync
SWOT Analysis
Let's be honest about where MobileMuuzaji stands right now — what it has going for it, where it's lacking, and what the landscape looks like.
Strengths
Built for the offline reality of its target market. A significant portion of Kenya's small business activity happens in areas with inconsistent mobile data coverage — markets, roadside stalls, semi-rural areas. Most inventory apps assume a reliable connection. MobileMuuzaji doesn't. The offline-first architecture means a market vendor in Gikomba or a small shop in Kisumu can still record every sale and track every product even when their data runs out or the signal drops.
Low barrier to entry. The app targets Android, which dominates the Kenyan smartphone market by a wide margin. There's no web portal to navigate, no desktop software to install — it's a single app that lives on the device the user already has.
Role-based access makes it suitable for small teams. A shop owner who employs a few staff members can give them access to record sales and manage inventory without giving them full administrative control. This is a meaningful feature for the target user who often delegates day-to-day operations to employees.
Flexible unit system. Supporting both weight and volume units might seem like a small detail, but it matters enormously in a market context where vendors sell everything from rice by the kilogram to cooking oil by the litre. A generic unit field wouldn't capture this accurately.
ACID-compliant transactions on the backend. Sales are processed as atomic transactions with row-level locking, meaning stock levels stay accurate even under concurrent load. This is the kind of backend reliability that many small-business tools at this price point (free) don't bother with.
Weaknesses
No financial reporting beyond basic earnings. Right now the app can tell you total earnings per product, but it can't generate a profit/loss statement, show trends over time with a chart, or export data to a spreadsheet. For a business owner trying to make informed decisions, this is a meaningful gap.
No price per unit tracking for sales. Sales currently record earnings as quantity × cost per unit, but there's no way to record a selling price that differs from the cost price. A vendor who buys onions at 50 KES/kg and sells at 80 KES/kg can't currently capture their actual profit margin — only their revenue.
No payment method tracking. M-Pesa, cash, and credit are all treated the same. In Kenya especially, distinguishing between cash and M-Pesa transactions is genuinely useful for reconciliation.
Employee management is online-only. Adding or removing employees requires an internet connection. For admin users in low-connectivity areas, this could be a frustrating limitation.
No data export. There's currently no way to export sales or inventory data to CSV, Excel, or PDF. For users who do their own bookkeeping or share reports with accountants, this matters.
Opportunities
Kenya's informal economy is enormous and underserved by software. The informal sector accounts for a significant share of Kenya's GDP and employment. The vast majority of these businesses have no digital tool for inventory or sales tracking — they rely on notebooks, memory, or nothing at all. The addressable market is large and the competition at the entry level is thin.
M-Pesa integration potential. Kenya has one of the world's most advanced mobile money ecosystems. An integration with M-Pesa's API would allow MobileMuuzaji to automatically reconcile sales against payments, which would be a genuinely compelling feature for Kenyan vendors and a strong differentiator against generic inventory apps.
Expansion to East Africa. The connectivity challenges and informal economy characteristics that make MobileMuuzaji relevant in Kenya are shared across Tanzania, Uganda, Rwanda, and Ethiopia. The offline-first architecture that was built out of necessity for the Kenyan market becomes a strategic asset if the app expands regionally.
WhatsApp Business integration. Many Kenyan small business owners use WhatsApp as their primary customer communication channel. A feature that lets users share stock availability or sales receipts directly via WhatsApp could meaningfully increase daily active usage.
Low-cost data analytics as a premium tier. As the user base grows, aggregated (anonymised) sales data could inform a premium analytics dashboard — showing seasonal trends, popular products by region, and demand forecasting. This could form the basis of a freemium revenue model.
Threats
Generic inventory apps from global players. Apps like Zoho Inventory, Wave, and various POS systems are available on the Play Store and some have free tiers. While they're not built for the Kenyan informal market specifically, they have significantly larger feature sets and established brand recognition. MobileMuuzaji needs to compete on local relevance and simplicity, not features.
Smartphone upgrade cycles. The app targets Android, but older, lower-spec Android devices are common in the target market. Performance on low-end hardware needs ongoing attention to avoid a poor experience for users on devices with limited RAM.
Data privacy and trust. Kenyan business owners may be cautious about storing sensitive sales and inventory data on a third-party server, particularly smaller vendors who operate informally. Building trust — through transparency about data handling, local hosting options, and clear privacy terms — will be important for adoption.
Connectivity improvements cutting both ways. As Kenya's mobile data infrastructure improves (5G rollout, cheaper data bundles), the offline-first differentiator becomes less compelling. The app needs to continue adding value in ways that remain relevant regardless of connectivity.
Where do we go from here?
If you remember from my intro I mentioned that I wanted to have data visualization tools in my app. Currently, I don't, and as a data scientist to be, this fact really bothers me. So in the next version of MobileMuuzaji, that will be the first addition (not just data visualization but data management tools as well). The main data management tools I am targeting are:
- A data visualization feature allowing customers to see trends in how their different inventory is performing
- Data download features with support for csv, json and xlsx file formats making it easier to export data and work on it elsewhere
- Potentially (this is a big thing, I don't even know if I'll manage) a mobile-based Jupyter Notebook integration so data can be analyzed and manipulated directly from inside the app
An additional feature alongside the data management will be adding the ability to manage purchases as well and map out profits. This tool will be especially useful, since, as every business owner knows, what matters is not the gross but the profit. Possible integration with VAT management would be a bonus.
But as of now, my mum loves the application, so that's what she will get. It is still being used at a small scale and has not been fully deployed for widespread use, but when it does, I'll update whoever was bored enough to follow my story.
Check out the github here