I Build Intelligent Systems That Ship

Generative AI developer based in Midrand, South Africa, building RAG, data platforms, edge AI, and business applications. Full stack from sensor signal to executive dashboard. Delivered solo.

Engineering capabilities

IoT & Edge Systems
01

IoT & Edge Systems

From IoT telemetry to custom hardware solutions

Physical computing systems built to capture telemetry from real devices, move that data into usable pipelines, and create custom hardware solutions when off-the-shelf equipment is not enough.

Key Deliverables
  • IoT telemetry capture from sensors, machines, and edge devices
  • Custom mechanical systems (dispensers, conveyors, positioning rigs)
  • ESP32/Raspberry Pi edge device programming
  • GPIO control for motors, servos, sensors, cameras
  • Edge-to-cloud telemetry pipelines (AWS IoT Core, Azure IoT Hub)
ESP32
Raspberry Pi
GPIO
Stepper Motors
AWS IoT Core
Telemetry
Python
See examples
Agentic AI & Intelligent Automation
02

Agentic AI & Intelligent Automation

Autonomous systems that reason, decide, and act

Agentic workflows where Claude/GPT models orchestrate multi-step processes autonomously: analysis, decisions, actions, exceptions, and human-in-the-loop review.

Key Deliverables
  • Autonomous AI workflows (n8n, Make, custom orchestrators)
  • RAG pipelines with vector databases and semantic search
  • Multi-model orchestration (GPT-4, Claude, Azure OpenAI)
  • Conversational AI with persistent memory (Xlink-Sage pattern)
  • Document analysis and intelligent data extraction
n8n
Make
Claude API
OpenAI
Vector DBs
RAG
LangChain
Docker
Python
See examples
Computer Vision & Edge ML
03

Computer Vision & Edge ML

Teaching machines to see and understand

Models deployed on edge devices processing real-world inputs in real time, with computer vision, OCR validation, and exception handling connected to cloud workflows.

Key Deliverables
  • Custom object detection model training (Edge Impulse, TensorFlow)
  • Edge deployment (TFLite on Raspberry Pi, embedded devices)
  • OCR pipelines with intelligent routing and validation
  • Azure Computer Vision & Cognitive Services integration
  • Image capture systems (industrial cameras, night vision)
Edge Impulse
TensorFlow Lite
OpenCV
Azure Computer Vision
Raspberry Pi
Python
OCR
See examples
Data Engineering & Backend Systems
04

Data Engineering & Backend Systems

The infrastructure that makes everything else possible

Pipelines, warehouses, and backend systems that turn raw operational chaos into structured, trustworthy data across SQL ecosystems and modern ETL.

Key Deliverables
  • Data warehouse design and implementation (Azure Synapse)
  • ETL architecture (Talend, SSIS, Azure Data Factory)
  • Database optimization and query performance tuning
  • REST API development and backend services
  • Realtime streaming and event processing
SQL
PostgreSQL
Azure Synapse
Talend
SSIS
Azure Data Factory
REST APIs
Python
See examples
Enterprise Applications
05

Enterprise Applications

Fullstack business systems that teams actually use

Operational apps with RBAC, audits, offline capability, and integrations, built on Power Platform and React/Supabase.

Key Deliverables
  • Warehouse and inventory management systems
  • Custom PowerApps with SharePoint/SQL integration
  • React + TypeScript apps with Supabase backend
  • Offline-first PWAs with barcode/QR scanning
  • Role-based access control and row-level security
PowerApps
Power Automate
React
TypeScript
Supabase
SharePoint
SQL
See examples
Analytics & Business Intelligence
06

Analytics & Business Intelligence

Operational visibility that drives decisions

30+ KPI dashboards, enterprise BI, automated reporting, and DAX/semantic models backed by solid engineering.

Key Deliverables
  • Enterprise Power BI implementations with advanced DAX
  • Realtime operational dashboards with live data
  • Automated report generation and distribution
  • Data modeling and semantic layer design
  • QlikView to Power BI migration
Power BI
DAX
SSRS
QlikView
Tableau
Data Modeling
Azure Synapse
See examples

Auditable SIM-Card OCR Pipeline

Case Study
Problem

Returned devices contained active SIM cards, but reading and recording printed SIM identifiers manually was slow, error-prone, and difficult to scale.

Results
Fine-tuned OCREdge captureEvidence-linked review
Raspberry Pi
ESP32
PaddleOCR
Kimi K3
AWS IoT
React

The Rare Profile

Most consultants fall into silos: AI specialists who've never touched hardware, IoT developers who can't build cloud backends, and data engineers who've never automated a physical process.

I'm the rare convergence, someone who can instrument the device, collect the telemetry, design the custom hardware, build the cloud pipeline, and deliver the executive dashboard. All as one person.

Accounting→Call Center→Business Analysis→SQL→Power BI→Azure→AI→Edge Computing→Hardware Automation
AI/MLHardwareCloudMErare convergence

Engagement Models

Project-Based

Fixed scope, fixed timeline, fixed price. Best for automation challenges, application builds, and system migrations.

Discovery & Architecture

Assess current state, identify opportunities, and deliver a technical roadmap with architecture recommendations.

Retained Advisory

Ongoing technical guidance for architecture reviews, vendor evaluations, mentorship, and strategic input.

Ready to Build Something Intelligent?

Whether it is a proof of concept, a production system, or a technical cofounder engagement, this is where it starts.

Ready to Talk?

Practical discussion about your goals and whether I'm the right person to help.

Recent Work

Case studies, demos, and deep dives on shipped projects.

Explore the Lab
4DAnalytics

© 2026 4D Analytics | Isaacson Shoko

Generative AI Developer | Midrand, South Africa

Privacy and analytics

This site records cookie-free interaction events on the production domain to distinguish genuine engagement from infrastructure traffic. It respects Do Not Track and Global Privacy Control and does not include names, email addresses, form messages, or assistant question text in analytics.