Engineering Lab:
Interactive AI & Predictive Models
Test our proprietary algorithms directly in your browser. Below you can interact with our Live Gemini RAG Chatbot, simulate traffic growth with our Predictive Time-Series Forecaster, and inspect our Multi-Touch Marketing Attribution Model.
Enterprise RAG (Retrieval-Augmented Generation) Chatbot
This live interactive widget demonstrates our custom RAG engine extracted from our production Python repository. Type any inquiry below or click a suggested prompt to watch the retrieval algorithm fetch verified facts from the knowledge base in real time:
Type a query or click one of the suggested chips above to test the live semantic context retrieval.
Predictive Website Traffic & Revenue Forecaster
Extracted directly from our Blog 43 (R / Shiny Analytics Project). Adjust the forecast horizon and growth scenarios below to see real-time time-series projections, visitor counts, and revenue impact:
Multi-Touch Attribution: Markov Chains & Shapley Values
Rule-based attribution models (First-Click, Last-Click, Linear) severely distort advertising budgets by failing to account for cross-channel assistance. Our model implements First-Order Markov removal probabilities and Shapley cooperative game theory:
First-Order Markov Chain Transitions
Models user click journeys as a directed state graph. The algorithm removes each channel one at a time and measures the resulting drop in overall conversion rate:
Shapley Value Cooperative Game Theory
Originating in cooperative game theory, channels act as coalition players. Evaluates the marginal contribution of each channel across all 2ⁿ channel combinations:
