01
Case Study ↓
// Phase 01Q3 2025
Ignite Literacy
// MAU · Feb 26156 / 175
89.1%
Copilot adoption
// Engagementvs benchmark
2.4×
LinkedIn Learning · 6 months sustained
I help organisations actually adopt AI — teaching the people, building the pipelines, shipping the use cases.
Contact MeI'm an AI enablement specialist with both the technical chops and the teaching instinct. I build agentic pipelines on n8n, Airia, Workato, Power Automate and Claude Code on one hand — and I run capability uplift sessions for non-technical staff on the other. The job is the same either way: turn AI from a buzzword into a habit. Let's build something incredible together.
Capability uplift sessions, lunch & learns, prompt-engineering bootcamps and exec briefings. I teach AI to people who don't code — HR, marketing, finance, operations — and to people who do.
I sit with business teams to surface where AI moves a real number, then run the pilot end-to-end. 30+ use cases identified at VRG, 10+ in active POC, 6 in pilot — structured pipeline, not pilot theatre.
Building automation that wires LLMs into the boring parts of work using n8n, Airia, Workato, Power Automate, Power Platform and Claude Code. Document intelligence, knowledge management, process orchestration.
Predictive models that earn their keep. Session-level revenue forecasting (R² 0.77), item-class forecasting up to R² 0.91, market-basket analysis, and Streamlit dashboards business teams actually use.
Reports, briefings, video, scriptwriting. Half my job is making technical reality readable for executives and operational reality readable for technologists. Both audiences are tired of jargon.
Literacy before tooling. Business ownership before build. Stage-gate every use case — idea → feasibility → POC → pilot → handover. Build alongside the future owner, not for them. Every solution ships with someone other than me who can extend it. Measure cycle time and rework, not training completion. Sustained cadence beats workshop theatre.
BCG says 74% of organisations never get AI past pilot. Kyndryl says only 14% successfully align workforce, technology, and goals. We chose the 14%. Literacy before technology. Culture before deployment. Business ownership before build. Governance before scale.
Self-hosted Reddit intelligence platform with a 3D globe visualising discussion activity across 125 countries in real time. Daily AI-generated city-level reports delivered to subscribers via email. Built and ran end-to-end — infrastructure, scraping pipeline, front-end, and content generation. Shut down April 2026 after a Reddit-level IP block; the technical architecture held up.
Dashboard for tracking AI upskilling across an organisation. Points engine, streaks, badges, and leaderboards sit on top of an activity heatmap and per-department progress charts. Designed to make adoption visible — something I needed at VRG before we had proper tooling for it.
RMIT industry capstone. F&B revenue forecasting for cinema chains — R² 0.77 session-level, 0.91 item-class on held-out data. Multi-output demand forecasting across hundreds of product segments, feeding directly into a concessions procurement model the ops team could act on.
Final-year project: an ML web app helping smallholder farmers decide what to grow and how to treat soil. Dual models — crop recommender from soil composition and a fertilizer recommender with confidence scores. XGBoost + Random Forest + SVM ensemble wrapped in a React front-end and Flask API.
Full-stack e-commerce platform for a pharmacy in Indore. Product catalogue, cart, Razorpay checkout, inventory management, and user auth — digitising a local medical retailer that had no online presence. Built collaboratively as a freelance project.