Eight interactive demonstrations spanning data processing, machine learning and generative AI — each exposing the full pipeline from input to result.
Live demonstrations bridge the gap between theory and practice by letting visitors interact with working data science systems. Each demo presents a complete pipeline — from raw input through processing to visualised results — proving production-ready capability rather than academic abstraction. The three categories span different complexity levels and domains, so visitors at any technical level find something relevant.
These tools make the critical but unglamorous work of cleaning, transformation and preparation both visible and interactive — turning a mysterious black box into a comprehensible process.
Upload messy CSVs; get analysis-ready data with every transformation explained.
Restructure, validate and document nested JSON from APIs, NoSQL and web scraping.
Visual construction for learners; execution-plan analysis for experts — one tool, two audiences.
Interactive explorations that make abstract algorithms tangible — adjust parameters, watch decision boundaries shift, and build intuition that mathematical explanations alone cannot convey.
Train multiple algorithms simultaneously on the same dataset; compare trade-offs at a glance.
Watch the bias–variance trade-off unfold in real time as you drag sliders.
Confusion-matrix centred analysis that turns abstract metrics into actionable decisions.
Cutting-edge GenAI applications with their underlying mechanics exposed — making transformer-era techniques accessible without requiring knowledge of attention mechanisms.
Experiment with prompting techniques side by side; see exactly how phrasing shapes model output.
End-to-end retrieval-augmented generation with every pipeline stage visible, not hidden behind progress bars.
| Demo | Category | What You Learn |
|---|---|---|
| CSV Data Cleaner & Validator | Data Processing | Outlier detection, imputation strategies, and audit trails for messy data |
| JSON Formatter & Schema Validator | Data Processing | Restructuring, schema validation, and schema evolution for nested data |
| SQL Query Builder & Optimiser | Data Processing | Visual query construction and execution-plan optimisation |
| Model Comparison Playground | Machine Learning | Algorithm trade-offs across accuracy, interpretability, and training time |
| Hyperparameter Tuning Visualiser | Machine Learning | Parameter effects, interactions, and Bayesian optimisation |
| Performance Metrics Explorer | Machine Learning | Threshold trade-offs and cost-sensitive metric selection |
| Prompt Engineering Laboratory | Generative AI | Few-shot prompting, chain-of-thought, and temperature effects |
| RAG System Demonstration | Generative AI | Retrieval pipelines and grounding generation in documents |
Key Insight — Interactive exploration accelerates learning beyond passive reading: visual feedback transforms complex processes into comprehensible systems, building intuition about algorithm behaviour and trade-offs that mathematical explanations alone cannot convey.
These tools are fully interactive in the DataScienceDNA app. Explore the Lab →