← All projects 06 · BI · ML LinkedIn Profiling - DataRaise Final-year project: analysis and forecasting of DataRaise's LinkedIn performance. Full data architecture (ETL, warehouse, BI) and follower-forecasting models (SARIMAX, regressions; R² = 0.54). PDF/TXT/JSON→Talend→PostgreSQL→Python ML→Flask→Power BI Talend · PostgreSQL · Python · Flask · Power BIESPRIT capstone FIG. 01Full data architecture (capstone)ETL chain → PostgreSQL warehouse → Python models → Flask & Power BI reporting. FIG. 02ML model comparison (R²)Evaluation of the forecasting models; best R² = 0.54. FIG. 03Follower-growth forecastProjection of the LinkedIn follower count (SARIMAX, regressions). ← Back to projects
