AMR.ALFAYOUMY
← HOME / PROJECTS / MOBILY AI/ML MODERNIZATION
ACTIVE PROGRAM · 2025-Present · Enterprise modernization

Mobily AI/ML Modernization.

Data Scientist & AI/ML Engineer
Cairo, Egypt
Source: resume + original portfolio + LinkedIn

An end-to-end modernization effort migrating 15 mission-critical telecom use cases from legacy SAS environments into Dataiku-centered ML and data pipelines.

Use cases
15
Mission-critical telecom workloads
Subscribers
13M+
Production business scale
Domains
4+
Core business areas
// 02 — CHALLENGE
WHY IT MATTERED

The hardest part was preserving business logic parity while also improving maintainability and performance.

Legacy analytics programs often embed years of assumptions, edge cases, and operational shortcuts that cannot simply be rewritten from scratch without careful validation.

// 03 — APPROACH
HOW I BUILT IT

I worked on modularizing flows into clearer data-preparation, feature, and inference layers so the system was easier to reason about and operate.

Where possible, logic was shifted to scalable Spark and SQL push-down patterns while validation frameworks compared legacy outputs with new outputs.

The work spans graph-style social network analytics, churn models, device risk, and commercial optimization use cases.

// 04 — HIGHLIGHTS
KEY TAKEAWAYS
  • Re-architecting 15 mission-critical telecom use cases
  • Supporting a subscriber base of roughly 13 million
  • Building distributed pipelines on Cloudera, Teradata, Spark, and Kubernetes
  • Covering churn, propensity, SNA, credit risk, and commercial recommendation systems
// 05 — OUTCOMES
RESULTS AND LESSONS
  • Helped establish a path from brittle legacy workflows to a more maintainable, governed ML platform.
  • Improved operational clarity by decoupling workloads into modular Dataiku flow zones and distributed processing layers.
  • Set up migration patterns that can support a broader modern MLOps lifecycle.
// 06 — STACK
THE TOOLS
Platform
Dataiku DSSSASClouderaTeradata
Processing
SparkSQLKubernetes
Use cases
ChurnPropensitySNACredit risk