@satishsaka
Building explainable, risk-aware decision systems for paid media capital.
I design and build decision-support systems for high-variance environments.
Currently developing MDU Engine, an explainable, risk-aware system for paid media capital allocation. My interests include decision logic, validation gates, confidence-aware outputs, auditability, and human-in-the-loop AI.
Writing here to document system design choices and lessons learned while building production decision infrastructure.
Nothing here yet.
Jan 6 · 2 min read · If you’ve worked with paid media long enough, you’ve probably felt this tension: You have data.But you’re not fully confident it’s telling the truth. Yet dashboards keep nudging you: “Increase budget.”“Performance is up.”“Momentum detected.” And qu...
Join discussion
Nov 14, 2025 · 4 min read · Performance marketing is in a strange place right now. Every platform is “smart.”Every dashboard is “automated.”Every tool claims “AI-powered optimization.” But spend a week inside any real campaign—finance, mobility, retail, SaaS, doesn’t matter—and...
Join discussion
Oct 12, 2025 · 3 min read · When browsers started killing cookies, most performance marketers panicked.When ad platforms increased automation and reduced visibility, most teams… trusted the black box. I didn’t. I’ve spent years running high-spend campaigns for different industr...
Join discussionAug 28, 2025 · 3 min read · When people talk about decision systems, they usually talk about outputs. Scale. Reduce. Optimise. But after working with real advertising data for several years, I’ve learned that the most important behaviour of a decision system often happens befor...
Join discussion