Marketing Technology × AI — TU Munich MSc (final semester) · Munich, Germany
I kept asking: what if every marketing decision could be formalized as an AI problem?
So I built the answer — one technique at a time.
A unified intelligence platform where each project solves a real marketing problem using a classical AI technique from my TUM curriculum. Not academic demos — tools real teams would use.
PLAN (CSP) → CREATE (ContentEngine) → ANALYZE (Bayesian + Journey) → OPTIMIZE (MDP planned)
↑
INTELLIGENCE (Competitor Intel)
↓
COMPLIANCE (Logic Engine planned)
| Project | Problem | AI Technique | Stack |
|---|---|---|---|
| 🧠 Journey Intelligence Engine | Which steps kill conversion — and why? | Markov chains · Value iteration · Absorbing chain | React 19 · D3 · TypeScript · Vercel |
| 🔍 Competitor Intel Monitor | Turn competitor pain into content opportunities | Multi-source signal pipeline · LLM classification | React · TypeScript · Claude API · Netlify |
| 📦 Return Root Cause Diagnosis | Why did the customer return? (without asking them) | Bayesian Networks · Noisy-OR · Backward inference | Python · NumPy · Streamlit |
| 📅 CSP Campaign Planner | Build a quarterly campaign calendar that satisfies 14 constraints | Constraint Satisfaction · CP-SAT · Backtracking | Python · OR-Tools · Plotly |
| ✍️ ContentEngine AI | Raw signal → 5 channel-native content drafts in <60 seconds | Structured prompt chaining · Batch pipeline | React · TypeScript · Vite · Vercel |
- MSc Management & Technology · TU Munich — Marketing major, CS minor
- BA Business Admin & Law · Koç University — IP specialization
- Infineon Technologies (2024–2025) — Marketing Analyst: AI agents, GA4/GTM, Tableau/Looker dashboards, automation workflows
- IUS Startup (2024–2025) — Co-Founder: AI-based IP monitoring platform
- Languages: English C2 · German B2 · Turkish C2
Law (IP / GDPR / EU AI Act) × AI Foundations (CSP, Bayes, HMM, MDP, Logic) × Marketing Operations
This combination is what I keep finding useful — and hard to find elsewhere.
- Project 6: MDP Optimal Contact Policy — state-dependent email cadence via value iteration
- Project 7: Logic-Based Compliance Engine — GDPR + EU AI Act as propositional logic + forward chaining
📬 LinkedIn · 📂 Notion Portfolio