Agentic & Generative AI
AI agents, RAG pipelines, vector retrieval, and grounded access to structured and unstructured enterprise data.
Hello, I'm Ghazali Akmal Rabbani
I build intelligent systems that turn data into business impact.
I design and lead Agentic AI, recommendation systems, forecasting, and end-to-end data products that make complex intelligence useful.
Ghazali Akmal RabbaniSenior Data Scientist
Core focusAgentic AI
4recommendation models delivered for Vision+
50%faster processing for a personalization algorithm
3.85MBA GPA · Magna Cum Laude
Core expertise
Technical depth, connected to business reality.
My work covers the complete path from opportunity framing and architecture to model delivery, testing, integration, and adoption.
Recommendation Systems
Behavioral, content-based, trending, seasonal, and personalization engines designed for real product environments.
Forecasting & Analytics
Forecasting systems, anomaly detection, testing, and decision-ready analytics translated into business action.
Technical Leadership
Cross-functional delivery, architecture, project planning, client consultation, and end-to-end quality assurance.
Selected work
Systems built for real decisions.
Vision+ · Media & Entertainment
Multi-Model Recommendation Engine
A portfolio of recommendation models designed to make content discovery more timely, personal, and relevant.
All Data International · Generative AI
Enterprise Knowledge AI Agent
An agent architecture that retrieves answers from documents, tabular files, and SQL sources through one conversational interface.
Enterprise Client · Industrial Analytics
Commodity & Operational Forecasting
Forecasting models for commodity, marine, and weather variables designed around operational decision needs.
Career perspective
From model building to leading the room.
Across media, analytics consulting, and enterprise AI, I have led teams through the practical details that determine whether a model becomes a useful product.
My MBA and technical background help me bridge business priorities, architecture choices, and delivery standards without losing sight of the people who will use the system.
More about my approachLet's build what matters