Sicelwesihle Myeza
Computer Science → Data → Systems

I build with code.
I work with data.
I'm learning how systems move.

Computer Science graduate from the University of the Witwatersrand and a Data Analyst at iqbusiness, working across data analysis, Python, SQL and business intelligence. Mathematics, probability and systematic thinking are the direction I'm steadily building toward.

Johannesburg, South Africa

View selected work → About me
LinkedIn ↗ GitHub ↗ Email ↗ Call ↗ Download CV ↗
Portrait of Sicelwesihle Myeza

Sicelwesihle Myeza · Johannesburg

About

A Computer Scientist working in data, building toward markets.

I studied Computer Science at the University of the Witwatersrand, where I got hooked on the idea that complex systems, whether software, data or eventually markets, can be understood by breaking them down and modelling them. That instinct pulled me from writing programs toward working with data.

Today I'm a Data Analyst at iqbusiness, working with Python, SQL and business intelligence tools to turn messy data into things people can actually decide on. Alongside the day job, I read economics and follow financial markets closely, and I'm drawn to the parts where mathematics, probability and systematic decision-making meet uncertainty.

That's the direction I'm deliberately building toward: quantitative. Not as a title I'm claiming today, but as the place all of this (computer science, data, statistics and markets) is pointing.

Beyond the work interests

Football
Strategy, systems, competition.
Chess
Patterns, calculation, patience.
Tennis
Adaptation, momentum, individual performance.
Markets
Probability, uncertainty, decision-making.
Experience

Experience

Data AnalystPresent
iqbusiness · Johannesburg
  • Clean and transform datasets and build ETL workflows for analysis-ready data.
  • Run exploratory analysis in Python and SQL, and maintain data-quality checks.
  • Build KPI dashboards in Power BI and communicate findings to stakeholders.
  • Collaborate in an Agile / Scrum delivery team.
Data AnalystJan 2025 – Nov 2025
University of the Witwatersrand
  • Analysed and visualised research datasets to support academic projects.
  • Applied Python and SQL to clean, join and summarise data for reporting.
Technical Lab AssistantMar 2022 – Nov 2023
Wits Mathematical Sciences Laboratory
  • Provided technical support and software installation across 200+ lab computers.
  • Improved workflows and system reliability for students and staff.
Education

Education

University of the Witwatersrand

BSc Computer Science

2020 – 2024 · Johannesburg, South Africa
Majors Computer Science & Computer Applications
Highlight First-class marks in mathematics (probability, statistics, linear algebra).
Relevant coursework
Advanced Analysis of Algorithms Machine Learning Database Fundamentals Software Design Operating Systems & Systems Programming Computer Graphics & Visualization Parallel Computing Computer Networks
Selected work

Selected work

Projects built in the open, where code and data answer real questions. Every link goes to a live repository.

01

V8: Quantitative Football Model in development

A production-style match-prediction and staking engine. Goal outcomes are modelled with Dixon-Coles (bivariate Poisson) using Bayesian shrinkage and time decay; probabilities are calibrated (isotonic / Platt, reliability diagrams, Brier / log-loss); bets are sized with fractional Kelly; and any edge is proven honestly through walk-forward backtesting and closing-line value. Every component maps to a transferable quant / risk technique.

PythonDixon-ColesCalibrationKelly stakingBacktesting
GitHub ↗
02

Credit Default Scorecard in development

An interpretable probability-of-default scorecard built the way a bank builds one: WOE / IV feature engineering, a logistic PD model, full validation (KS, Gini / AUC, PSI stability), PDO score scaling with reason codes, and a gradient-boosted benchmark to quantify the interpretability trade-off. Runs on the German Credit dataset and deliberately avoids leakage-prone sources.

PythonCredit riskWOE / IVLogistic PDValidation
GitHub ↗
03

Survey Overlap QC in development

Quality-control tooling for survey datasets, flagging overlap and duplication so responses are clean and analysis-ready before any modelling. The kind of unglamorous data-integrity work reliable analysis depends on.

PythonData qualitySurveys
GitHub ↗
04

Survey Compact-to-Binary in development

A transformation utility that reshapes survey responses from a compact, multi-select encoding into a tidy binary / indicator matrix: the format analysis and modelling actually need.

PythonData transformationSurveys
GitHub ↗
Skills

Programming

PythonJavaC++JavaScriptHTML / CSS

Data

SQLPandasNumPyMatplotlibSeaborn

Business Intelligence

Power BITableauExcel

Engineering

Git / GitHubAPIsWeb ScrapingJupyterETL
Trajectory
01

Computer Science

Foundation

BSc Computer Science, University of the Witwatersrand. The systems-thinking everything else is built on.

AlgorithmsMachine LearningSystemsSoftware Design
02

Data

Current chapter

Data Analyst at iqbusiness, using Python, SQL and BI to turn raw data into decisions.

PythonSQLETLPower BIEDA
03

Markets

Building toward

Quantitative finance and systematic strategy: the long-term direction I'm studying my way into.

Financial MathsStatisticsTime SeriesSystematic
Currently exploring

What I'm learning next, and the questions driving it.

The topics I'm actively studying as I move toward quantitative finance, and the research questions I find worth chasing.

Statistics Financial Mathematics Time Series Optimisation Systematic Strategies Algorithmic Trading

// These are directions I'm learning toward, not completed projects or claimed results.

Research interest

Market regimes

How can statistical techniques identify different market environments, and adapt to them?

Time series / Statistics
Research interest

Mean reversion

Can temporary deviations from an estimated equilibrium provide useful, testable signals?

Statistics / Backtesting
Research interest

Portfolio optimisation

How can mathematical optimisation balance expected return against risk?

Linear algebra / Optimisation
Research interest

Algorithmic trading

How can systematic rules be translated into robust, testable strategies?

Python / Markets / Software
The path

A Computer Scientist building his way toward quantitative finance.

Foundation
Computer Science
University of the Witwatersrand
Then
Software & Data
Wits · research data analysis
Now
Data Analytics
iqbusiness
Building toward
Quantitative Finance
Future direction
Long-term
Systematic / Algorithmic Trading
Long-term ambition
Contact

Let's build something meaningful.

Open to data and analytics roles in South Africa, and always happy to talk about markets, maths and systematic thinking.

Emailsicelwesihlemyeza@gmail.com Call+27 62 162 8387 LinkedInin/sicelwesihle-myeza ↗ GitHub@esiihle ↗ CVDownload PDF ↗