// skills & expertise

40+ years building, breaking, and rebuilding machines — from discrete logic gates to large language models.

LLM & Agentic AI

Engineering with large language models — from API integration through fully autonomous multi-agent pipelines.

🤖

LLM Engineering

System prompting, tool use, RAG pipelines, context window management, multi-turn conversation design. Daily driver: Claude Sonnet/Opus via API and Claude Code.

Agentic Systems

Multi-agent orchestration, autonomous code generation and execution, Claude Code in YOLO mode, agent memory, CLI integration for compile-debug-deploy loops.

🔗

Model Context Protocol (MCP)

Building and consuming MCP servers to connect LLMs to live data sources, databases, APIs, and local tools.

Machine Learning & Deep Learning

Classical ML through modern transformer architectures — built, trained, and deployed at production scale.

🧠

Deep Learning

Transformer architectures, CNNs, RNNs/LSTMs, attention mechanisms, fine-tuning, LoRA/QLoRA PEFT. Built and trained models from scratch before it was cool.

📊

Classical ML & Statistics

Feature engineering, ensemble methods, Bayesian inference, time-series analysis. The fundamentals still matter when the data is small or interpretability is required.

🗣️

Natural Language Processing

Clinical NLP, named entity recognition, information extraction from medical texts, text classification, embedding-based search.

Medical AI

The intersection of clinical medicine and machine intelligence — where domain knowledge is the competitive moat.

🏥

Clinical Decision Support

ML models for diagnostic assistance, risk stratification, and outcome prediction. Built systems that operate in low-resource clinical settings.

📡

Medical Data Engineering

EHR/EMR pipeline development, DICOM image processing, anonymisation, and dataset curation for downstream ML.

Engineering & Systems

Hardware, firmware, embedded systems — the physical layer that makes software real.

🔧

Electronics & Embedded Systems

From discrete logic gates (1975) to modern MCUs. Arduino, ESP32/ESP-IDF, Raspberry Pi, IoT sensor networks, HAM radio (NE4AL), LoRa-based disaster comms (DiMSAR Kenya).

🐍

Python & Software Engineering

Primary development language for AI/ML work. Flask, FastAPI, async Python, packaging, testing. Drops to C/C++ when performance demands it.

📈

Algorithmic Trading

ML-based trading strategies, MQL5 development, neural networks for market prediction. Achieved personal financial independence via systematic trading.

Blockchain & Web3

Bitcoin since genesis (2009). Solidity smart contract development, Ethereum ecosystem, Cardano. Deep understanding of cryptographic primitives.

career_timeline.sh
1975 Built first computer from discrete components (Intel 8080 era)
1976 Started using EMACS; adopted Assembly → C → LISP
1990s C++, Fortran, Pascal; founded CyberCollege Africa (first MOOC on continent)
2000s Deep dive into ML, neural networks, algorithmic trading (MQL)
2009 Bitcoin genesis block; mined thousands of coins; built blockchain tools in Solidity
2015+ Python-first AI pipeline; PyTorch, TensorFlow, medical NLP, >600 AI practitioners trained
2024+ LLM engineering, agentic systems (Claude Code), multi-agent orchestration
now _