Gagandeep B.

I help brands and agencies work out which marketing spend drives incremental revenue, and where to move budget next. At an agency I've delivered Bayesian Marketing Mix Modeling (MMM) for 6+ clients across retail, hospitality and e-commerce. One engagement led to a $127K budget reallocation the client adopted. WHAT I DO • MMM build and validation: Google Meridian, PyMC-Marketing and Robyn. Adstock, saturation, ROI priors, uncertainty and budget optimization. • Independent MMM audits: data readiness, channel confounding, multicollinearity, prior sensitivity, holdout checks and model stability. • Incrementality testing: geo-lift, A/B tests, power analysis, causal impact, and using experiments to validate or calibrate MMM. • Reporting clients can act on: branded PowerPoint and Word deliverables that turn model output into clear commercial decisions. • Secondary: production LLM/RAG systems, Python/FastAPI services and BigQuery pipelines on Google Cloud. WHY WORK WITH ME I check every model's assumptions instead of trusting its output. I've caught and fixed errors, such as a missing intercept in a synthetic control model and overstated effects in CausalImpact, before they reached a client decision. I make uncertainty and identification limits visible, and I challenge ROI estimates that look too good. BEST FIRST STEP Need to know if your data is MMM-ready, get an independent review of an existing model, or design an incrementality test? Send me your business question and a sample of your data, and I'll propose a clear scope, deliverables and decision framework.