EaziMarketing — WhatsApp & AI Marketing SaaS
Designed and built a full-stack SaaS platform serving SMEs across South Africa and globally. EaziMarketing combines WhatsApp automation with AI-powered social media marketing, a built-in CRM, Google Ads AI campaign builder, and analytics — all in one dashboard. Built for contractors, medical practices, estate agents, and service businesses who need to automate lead capture and marketing without a full marketing team.
The platform handles lead responses in under 60 seconds, generates AI social content, manages multi-platform scheduling, and tracks ROI across channels — end-to-end, from concept to production.
Visit EaziMarketing.com ↗
SiteReady — AI Safety & Compliance Platform
Built a WhatsApp-first SaaS that generates OHSA-compliant safety files and tender documentation for South African contractors — in minutes, not days. Contractors simply message the AI on WhatsApp, answer a few questions, and receive professional PDFs covering 30+ document types directly on their phone.
The platform covers 18 industries, 225 verified legal requirements, and produces everything from safety files and risk assessments to CIDB tender packs and appointment letters — all legally compliant and AI-generated from domain-specific datasets.
Visit SiteReady.co.za ↗
Multi-Tenant eCommerce Analytics Platform for Amazon Sellers
Architected and led development of a multi-tenant SaaS analytics platform for Amazon marketplace sellers — featuring a WordPress-based client dashboard, Power BI and Tableau reporting, Azure cloud infrastructure, and automated data pipelines pulling from the Amazon Selling Partner API.
Secured $150K in Microsoft vendor grant credits for the production portal, significantly reducing infrastructure costs. Led a distributed development team, managed client relationships, and delivered complex SSIS ETL packages processing live Amazon API data into production SQL databases.
AI Metadata Extraction for Quasi-Government Archival Division
Modernised metadata extraction processes for a quasi-government archival organisation — replacing manual cataloguing with AI-driven Python pipelines that dramatically improved data accessibility and retrieval efficiency. Also developed AI RAG (Retrieval-Augmented Generation) applications integrating external APIs to streamline workflows and resolve chronic data consumption bottlenecks.
Additional work included streamlining ETL processes using regex-optimised pipelines, developing AI Agents using Generative AI for in-house projects, and building with Next.js, TypeScript, and Tailwind for modern frontend delivery.
B2B AI SaaS Architectures — Analytics, Agents & IoT
Designed and governed a range of B2B AI SaaS architectures including analytics platforms trained on domain-specific datasets, AI agents, IoT-integrated AI systems, and cloud-native data pipelines. Delivered custom AI chatbots integrated into client websites, trained on proprietary datasets including documents, PDFs, and URL data.
Also developed an on-premises ChatGPT-style data analysis bot in Python — running locally on client hardware and using NLP queries against custom datasets, enabling secure AI analytics without cloud dependency.
AI-Powered Guest Review Analysis Dashboard
Built an NLP analytics platform that processed over 1,200 Airbnb and Booking.com guest reviews — automatically surfacing complaint categories, sentiment trends, and per-property performance metrics in a single dashboard. Property managers went from manually reading reviews to having AI flag issues, track patterns over time, and generate downloadable reports instantly.
The dashboard segments reviews by property, date range, and complaint category — cleanliness, noise, check-in issues, location — with word cloud visualisation and at-a-glance KPIs showing total reviews analysed and average ratings across the portfolio.