JIMMY OKOTH 30%

Jimmy Okoth

Data engineering
& the road to 100%

Nairobi, Kenya. Building for the world.

30%

Building toward
100%

Current
Data Engineering
Destination
Data Architect
+ Tech Entrepreneur
Status
Still building.
EXPLORE THE JOURNEY ↓

01Introduction

I'm Jimmy. I trained in actuarial science and spent a while unsure where I fit. Then I started pulling on data systems and couldn't stop.

Now I'm learning how data moves between systems and why pipelines fail at the worst possible moment. I'm also learning what a business actually needs from its information, which is rarely what the dashboard shows.

I'd rather build something that works than describe something that sounds good. This site is the record of that. It will be wrong in places. That's part of the point.

03

Projects

02

Writing categories

30%

Current progress

02Currently building

Areas in motion.

These aren't proficiency scores. They show where each area sits in my learning right now, and they move as I do.

▬ Foundation

SQL

Window functions, query plans, SQL someone else can read.

Python

Pipelines, API clients, tests.

▬ Active focus

Data Engineering

Ingestion patterns, idempotent loads, orchestration.

Data Modeling

Kimball first, then knowing when to break it.

Cloud

Azure storage, Data Factory, cost and identity.

▬ Studying

Spark

PySpark on real volumes. Why jobs spill.

▬ Horizon

Distributed Systems

Consistency, failure, the papers everyone cites.

Architecture

Trade-offs across whole systems.

03Selected work

Projects told as problems, not features.

01In progress · 2026

REDDIT AUDI DATA PLATFORM

Turning a noisy public API into a warehouse I'd trust to answer questions.

An end-to-end pipeline that pulls posts from Audi-focused subreddits, lands them raw, cleans them in layers and models them for analysis.

PythonAzureSQLETLData Warehouse
VIEW CASE STUDY →LIVE DASHBOARD ↗GITHUB ↗
02In progress · 2026

NEWCOMER CREDIT DATA PLATFORM

Fair, point-in-time affordability features for people with little or no local credit history.

Python 3.12PostgreSQL 17dbt CoreMakeGit
  1. Project base and workstationDone
  2. Synthetic data generatorDone
  3. Loading the raw layerIn progress
  4. First dbt modelPlanned
  5. Affordability featuresPlanned
VIEW CASE STUDY →GITHUB ↗

04The journey

0 → 100

A map, not a measurement. Everything past the current marker is where I intend to go, not somewhere I've been. The destination is allowed to change.

  1. 00%ACTUARIAL SCIENCE Reached2022 · The starting point
  2. 30%DATA ENGINEERING CURRENTBuilding pipelines end to end, learning to model data properly and making things fail loudly instead of quietly.
  3. 40%CLOUD + PRODUCTION SYSTEMS Ahead
  4. 60%DISTRIBUTED SYSTEMS Ahead
  5. 80%SYSTEM DESIGN + ARCHITECTURE Ahead
  6. 100%DATA ARCHITECT + TECH ENTREPRENEURThe destination, for now

05Writing

Two notebooks. One for the systems, one for the person building them.

■ Technical / Build Log

How it works, and how it broke.

12 OCT 2026

My first Audi pipeline lied to me

I rebuilt my Reddit Audi data pipeline after my Azure subscription died. Rebuilding it exposed a bug that flipped one of my headline findings.

READ THE POST →

○ Life / Field Notes

What it costs, and what it’s for.

23 OCT 2026

Give Yourself a Mission to Live

READ THE POST →

I don't know exactly where this ends. But I know which direction I'm moving.

I'm building the technical ability, the judgment and the independence to one day build something of my own. This website is the record.

● 30% STILL BUILDING.

SAY HELLO →

or write to officejim727@gmail.com