Raed Mokdad
AI Systems Engineer

AI Systems &
Data Engineering

I build AWS/Python pipelines, search & recommendation, and RAG/embedding systems that stay reliable after launch.

Proof of production delivery: CarTic system (live since 2008) • PV rooftop pipeline shipped into company environment • recommender evaluated on 62k real-market listings.

View Work Get in Touch

Who I Am

I'm Raed Mokdad. I design, build, and operate data and GenAI systems on AWS, end to end, from ingestion and orchestration to deployment, monitoring, and continuous improvement. I bring deep engineering fundamentals to modern data & AI challenges, enabling reliable solutions under real-world conditions and well-understood trade-offs.

I think in outcomes before I think in tools. I start by understanding how your business works, your processes, constraints, data flows, and where time, cost, or quality is lost. From there I guide the key decisions and define a plan that turns your goal into a concrete system.

I work in two modes: as part of a team to accelerate implementation, or independently end-to-end with clear milestones, transparent communication, and systems teams can reliably operate.

What I Design and Deliver

1) Data Engineering on AWS (Python)

  • Multi-source ingestion (APIs/files/DB), cleansing, normalization
  • Storage & retrieval: S3, DynamoDB, OpenSearch
  • Data quality, monitoring, cost control, handover
Result: Reliable, observable pipelines and clean data your team can trust.

2) Search & Recommendation Systems

  • OpenSearch indexing, filtering, ranking
  • Hybrid retrieval: vector search + metadata filters
  • Evaluation & tuning (Hit@k, precision/recall), latency/cost optimization
Result: Intent-aware recommendations that feel right, driving better decisions so users stop searching and commit.

3) GenAI Integration (RAG & Embeddings)

  • RAG design (chunking, retrieval strategy, citations)
  • Embeddings strategy, caching, vector retrieval
  • Guardrails, validation, fallback logic, quality evaluation
Result: GenAI features with guardrails, evaluation, and predictable behavior in production.

4) System Integration & Automation

  • API integration, event-driven workflows, notifications
  • Orchestration with auditability and clear ownership
Result: Less manual work, fewer errors, faster turnaround.

How I Work

Clear communication. Results-focused.

Project-Based

Fixed scope, clear deliverables, defined timeline.

Embedded in Your Team

I join your team temporarily and ship alongside your developers (hands-on delivery, reviews, documentation, handover).

Advisory

Architecture reviews, technical assessments, and mentoring to align on key decisions before implementation.

Selected Work

Production since 2008

CarTic — Real-Time Vehicle Search & Communication Platform

API integration and data filtering with WhatsApp/VoIP automation for time-critical seller outreach. Live B2B platform since 2008, designed to stay reliable under frequently changing upstream constraints.

Real-time Processing Distributed Systems
View details
In use (Energietechnik West)

Solar Rooftop Identification System (PV Potential)

Geo-data pipeline to identify suitable PV rooftops and generate actionable leads. Shipped into a company environment with real-world data.

Data Pipeline Lead Scoring
View details
Evaluated on 62k listings

Recommendation System

Intent-aware recommendations using semantic retrieval + hard filters to surface convincing alternatives. Evaluated on 62k real-market listings with measurable relevance metrics (e.g., Hit@10).

OpenSearch Embeddings
View details
Demo (governed analytics)

Natural Language to SQL System (Talk to Data)

Built a "talk-to-data" interface (NL→SQL) with access control, safety checks, and auditability. Designed for predictable behavior and reduced risk in self-service analytics.

Prompt Engineering SQL Safety
View details
Demo (production-style)

Retail Sales Forecasting System

End-to-end demand forecasting pipeline on AWS, from ingestion and feature prep to model and dashboard. Built as a production-style demo on the Rossmann dataset (S3/Glue/Athena/SageMaker/QuickSight).

Feature Engineering Model Evaluation
View details
Blueprint (AWS analytics)

AWS Cloud Data Foundation

Reusable AWS serverless analytics foundation (S3/Glue/Athena, Bronze/Silver/Gold layers, quality checks, dashboards). Blueprint for standardized delivery across projects and teams.

Data Quality Cost-Optimized Architecture
View details

Get in Touch

Let's talk about your project. I typically reply within 24–48 hours with a recommended approach and next steps.

Prefer email? mokdad@raedmokdad.com

If you need someone who turns complex data into clarity and AI hype into working systems — let's talk.