Over the past six months, I embarked on an intensive cancer research project that evolved from broad conceptual exploration to sophisticated data engineering. The journey began with Grok and later accelerated dramatically with Claude AI combined with coding capabilities. 1. Research of Cancer through Grok (Months 1–4) • Opened a dedicated long-term project in Grok and systematically asked nearly 300 questions to understand the full scope of cancer — from molecular mechanisms, tumor heterogeneity, metastasis pathways, and the central dogma (DNA → RNA → Protein) disruptions, to clinical challenges in early detection and treatment resistance. • Explored key technologies including microfluidic biochips, single-cell separation, CTC isolation, immunofluorescence, NGS/PCR downstream analysis, and the differences between liquid and tissue biopsies. • Investigated global cancer epidemiology, existing detection platforms (e.g., CellSearch, CytoAurora CMS), and the limitations of current data sources. • Although Grok provided valuable structured explanations and technical depth, the process required extensive manual follow-up, repeated clarification, and cross-validation because cancer research involves highly interconnected, multi-disciplinary knowledge. 2. Data Research and Collection (Months 3–5) • Discovered that global cancer data is highly fragmented across sources (WHO, IARC, GLOBOCAN, national registries, academic papers). • Most datasets offered only country-level incidence, mortality, and prevalence, with inconsistent years (some limited to single years like 2002) and varying formats. • Collected raw data from multiple sources and categorized them into two tiers: • 1.0 Comprehensive Data: Full incidence, mortality, and prevalence figures with metadata. • 0.5 Percentage Data: Derived or estimated percentages generated or validated through Grok. • Performed extensive data cleaning to standardize country names, cancer types, and units. 3. Data Integration through Excel, Power Query, Power BI and DAX (Months 4–6)