KPI Visualization Done Right: Parsing Any Data Format in the Browser
How we built a multi-format data parser that handles CSV, JSON, XML, XLSX, JSONL, and YAML entirely in the browser — and auto-detects KPIs with statistical analysis.
Every business runs on KPIs. But getting to the KPIs usually involves exporting data, opening Excel, writing formulas, and building charts manually. We built KPI Master to eliminate that friction.
The Multi-Format Parsing Challenge
Supporting six different file formats — CSV, JSON, XML, XLSX, JSONL, and YAML — meant solving distinct parsing challenges for each. Papa Parse handles CSV edge cases. Native JSON.parse works for JSON/JSONL. DOMParser traverses XML. SheetJS handles Excel. js-yaml parses YAML. All running in the browser.
KPI Auto-Detection
Our engine uses three signals to detect KPIs: numeric ratio in columns, column name pattern matching, and variance analysis. For each detected metric, we compute sum, average, median, min, max, count, and standard deviation. For time-series data, we add growth rate and trend direction.
Rendering Without a Charting Library
Rather than bundle a heavy library, we render visualizations using inline SVG. Sparklines are simple polyline paths, and bar charts use SVG rect elements with gradient fills — lightweight and fast.