KG.Mokoena — Where Code Meets Curiosity
Kgopotso Mokwena:
AI/Software developer.
Building production AI systems and full-stack applications. From RAG pipelines to real-time delivery platforms — turning complex problems into elegant solutions.
Kgopotso Mokwena@lab — zsh — 80×24
The Journey
From Java to AI Systems
Not your typical path. Started with academic Java at UMP, self-taught full-stack JavaScript, then built production AI systems for real paying clients — all while completing an NQF7 degree and teaching 50+ students.
The Academic Foundation
Advanced Diploma in ICT (NQF7) from University of Mpumalanga, completed November 2025 and awarded May 2026. Also holds an NQF6 Diploma (awarded May 2024). As a Teaching Assistant, tutored 50+ students in Python and Java, improving average test scores by 21%. Teaching is the fastest way to expose what you don't actually know.
The Self-Taught Pivot
While peers stopped at tutorials, I shipped for clients. Self-taught Node.js, PostgreSQL, React Native, then the full AI/ML stack. Built Axiora AI — a multi-tenant RAG system with a 4-stage retrieval pipeline achieving 94% relevance accuracy on 10,000+ documents. IBM-certified in Data Science and Agile to back it up.
The Full-Stack Reality
From real-time food delivery with PostGIS geospatial tracking (Happy Deliveries) to AI travel concierges (Casalinga Tours) and vehicle verification platforms (Before You Sign). Two freelance clients. Production systems. Real business impact. Cross-disciplinary by default — because the best solutions rarely live in a single domain.
// My approach to software engineering
const philosophy = {
buildForScale: true,
automateEverything: true,
teachWhileLearning: true,
measureImpact: "94% RAG accuracy | 90.07% stock prediction | 70% pre-booking automation | 6 production projects",
currentStack: ["Python", "Node.js", "FastAPI", "PostgreSQL", "LLMs/RAG"]
};
Artifacts
Technical Projects
Production systems built for real clients and real problems — AI pipelines, delivery platforms, verification tools, and data models. Not tutorials. Not clones.
Arsenal
Technical Stack
Technologies I use to build production-ready systems. From databases to AI pipelines.
AI & LLMs
Data & Storage
Frameworks
Languages
The Path
Experience & Education
From student to teacher to builder. Every step taught me something about code, systems, and people.
Teaching Assistant — Application Development
Tutored 50+ students in Python and Java. Used automated grading and rubric-based assessment to reduce marking workload. Improved average test scores by 21% through supplementary lectures and hands-on debugging sessions.
Freelance Software Developer
Delivered two production AI systems for paying clients. For Casalinga Tours: an AI travel concierge powered by Google Gemini 2.5 Flash, automating 70% of pre-booking inquiries. For Before You Sign: a vehicle verification platform with QR-based audit trails, cryptographic signing, and a 7-step verification workflow.
Advanced Diploma in ICT (Applications Development) — NQF 7
Completed November 2025, formally awarded May 2026. Also holds Diploma in ICT (NQF6), awarded May 2024. Coursework included Advanced Database, Emerging Technologies, Cybersecurity, Human Computer Interaction, Research Methodology, and Project 400. Distinction in Research Methodology.
IBM SkillsBuild Certifications (×5)
Data Fundamentals (Dec 2025), Python 101 for Data Science (Feb 2026), Data Analysis with Python (Feb 2026), Applied Data Science with Python (Feb 2026), and Agile Explorer — Powered by Agile at IBM (Feb 2026). Five certifications across data science and Agile methodology.
Field Notes
Development Notes
Quick observations, experiments, and discoveries from the digital laboratory.
Current Focus
Workbench
What's currently on the workbench — active projects, experiments in progress, and things being built.