Engineering Intelligence
We solve real-world problems with software, data, and AI.
We start with the problem, understand it through data, and engineer the system that fixes it. Then we measure whether it worked.
Born in Kenya. Built in Africa. Engineered for the world.
Software · Data · AI
- Problem
- Data
- Intelligence
- Engineering
- Impact
We start with the problem, not the technology.
The problem decides the solution. The solution decides the technology.
We don't build things because we can. We build because something needs to be solved.
How we work
Bring us something messy. Here's where we start.
- 01
Understand the problem.
Who has it, what it costs them, and what better would look like.
- 02
Map the system.
The people, processes, data, and tools already involved.
- 03
Find the highest-leverage fix.
Sometimes that's software. Sometimes it's analytics, automation, or AI. Sometimes it's a change in process.
- 04
Build the smallest useful solution.
Something real, in use, as early as possible.
- 05
Measure the impact.
Against a number we agreed on before we started.
- 06
Improve.
What we learn goes into the next version.
Problems we work on
You might recognise one of these.
“Our operations run on spreadsheets, paper, and memory.”
Operations and business systemsCore today“People can't find or reach the opportunities that fit them.”
Education and access to opportunityCore today“We have data, but it doesn't help us decide anything.”
Data and decisionsDeveloping“We want to see what's coming, not only what happened.”
Prediction and intelligent systemsDirection
Software is the foundation. Intelligence is the direction.
Software
Systems that people and organisations use every day.
Core todayData
Systems that collect, organise, and explain information.
DevelopingIntelligence
Systems that learn, predict, and assist. Today: AI assistants built into our products. Next: models trained on real data.
Developing
Selected work
Real problems we've built for.
Education & Career Technology
CareerNext
Live at careernext.co.ke
- Problem
- Kenyan students have to make sense of KCSE results, KUCCPS choices, courses, and institutions, with the information scattered and the maths hard.
- Built
- A platform built to help students navigate all of it. It includes a cluster-points calculator, a course matcher across five pathways, and an AI guide grounded in each student's results.
Django · PostgreSQL · M-Pesa · GPT-4oRead the case study
Demonstration dataProperty Management Technology
LandlordPMS
In development
- Problem
- Landlords and property managers juggle tenants, rent, payments, and finances across SMS, notebooks, and WhatsApp.
- Built
- One phone-first system for all of it. It matches M-Pesa payments, tracks arrears, and works offline and over USSD.
Django · M-Pesa · USSD · ClaudeRead the case study
A technology company in the making.
MLLabz was started in Kenya by Meshack Limo, a Data Science student at JKUAT who began by learning to build software. We're growing deliberately: from software, into data, then machine learning and AI.
Stage 1
The Builder
We are here
Stage 2
The Data Company
Stage 3
The Intelligence Company
Stage 4
The Product Company
Stage 5
The Technology Enterprise
Have a problem worth solving?
Tell us the problem. Let’s engineer the solution.
Or email us at limomeshk@gmail.com