# SprkLogs

> SprkLogs is a free, local-first Windows desktop app that reduces large Apache Spark event logs and generates evidence-based technical analysis without sending the raw log directly to an LLM.

> SprkLogs e um aplicativo desktop para Windows, local-first, que reduz grandes logs de eventos do Apache Spark e gera analise tecnica baseada em evidencias sem enviar o log bruto diretamente para uma LLM.

## Summary
- Product: Windows desktop analyzer for Apache Spark event logs
- Audience: Data Engineers, Data Analysts, Data Architects, platform teams
- Core benefit: Reduce hundreds of MB or 1 GB+ logs into compact analytical context for precise diagnosis
- Privacy model: local reduction before any optional AI analysis
- Website: https://alexvalsechi.github.io/sprklogs/
- Source code: https://github.com/alexvalsechi/sprklogs
- Releases: https://github.com/alexvalsechi/sprklogs/releases

## Features
- Reduce large Spark UI logs into a compact technical report
- Identify bottlenecks, slow stages, task skew, and suspicious execution patterns
- Produce evidence-based recommendations tied to real job metrics
- Support hybrid architecture with local processing before optional LLM calls
- Export analysis for engineering workflows

## Use Cases
- Diagnose slow Spark jobs in production
- Review job execution with evidence before tuning partitions and shuffles
- Prepare a technical summary for pull requests, Confluence, or sprint optimization work
- Analyze logs too large to be sent directly to an LLM

## Technical Details
- Web: static marketing site hosted on GitHub Pages
- Desktop: Electron + TypeScript
- Backend: FastAPI + Python
- AI support: reduced report can be analyzed with supported LLM providers

## Links
- Main site: https://alexvalsechi.github.io/sprklogs/
- Full LLM context: https://alexvalsechi.github.io/sprklogs/llms-full.txt
- Portuguese LLM summary: https://alexvalsechi.github.io/sprklogs/llms-pt.txt
- Portuguese full context: https://alexvalsechi.github.io/sprklogs/llms-full-pt.txt
- GitHub repository: https://github.com/alexvalsechi/sprklogs
- Releases: https://github.com/alexvalsechi/sprklogs/releases

## Portuguese Notes
- Idioma principal para SEO e marketing: pt-BR
- O publico principal inclui engenheiros de dados, analistas de dados e arquitetos de dados no Brasil
- Para contexto completo em portugues, prefira: https://alexvalsechi.github.io/sprklogs/llms-full-pt.txt