I build reliable data collection, transformation, scalable vector retrieval, and delivery systems.
Data & AI systems engineer specialized in designing scalable ETL/ELT pipelines, analytics modeling, and low-latency RAG vector architectures. My work spans robust web data extraction, analytics transformation with dbt, high-speed API development, and deterministic clinical decision-support systems. I prioritize tested code, sub-second latency, clear documentation, and dependable data products.
| Area | Technologies and Practices |
|---|---|
| Data Engineering | ETL/ELT, pipeline architecture, data validation, automated extraction, event processing |
| RAG & Vector Retrieval | Hybrid dense/sparse search (BGE ONNX + BM25), Groq LPU inference, anti-hallucination guardrails |
| Analytics Engineering | dbt, SQL, dimensional modeling, transformation testing, data quality metrics |
| Backend & API Systems | Python, FastAPI, REST APIs, Web Speech AI, async task processing |
| Data Platforms & Storage | Snowflake, PostgreSQL, Vector Stores, Redis |
| DevOps & Automation | Docker, Git, GitHub Actions, CI/CD pipelines, automated test suites |
An evidence-grounded clinical AI decision-support platform designed for breast oncology, strictly adhering to UK NICE guidelines:
-
Hybrid Retrieval Engine: Combines FastEmbed ONNX dense embeddings with Rank-BM25 lexical search (
$0.65\text{ Dense} + 0.35\text{ BM25}$ ) under$<50\text{MB}$ RAM. -
Deterministic Confidence Gate: Hard similarity threshold (
$S \ge 0.58$ ) preventing hallucinations on out-of-domain medical queries. - Multimodal Interface: Synchronized PDF evidence preview drawer, real-time speech dictation, and speech synthesis voice output.
Python · FastAPI · Groq LPU · FastEmbed ONNX · BM25 · Docker · Speech AI
dbt models for transforming, testing, and modeling Earnest platform data for analytical queries.
dbt · Data Modeling · Testing · Analytics Engineering
A documented collection of 39 Python and JavaScript scrapers for Egyptian retailers, including cleaned sample datasets.
Python · JavaScript · Web Scraping · Data Extraction
A Python project exploring backend integration between FastAPI and Snowflake data warehousing.
Python · FastAPI · Snowflake · Async REST
- Designing high-throughput, reproducible ingestion and transformation workflows
- Building memory-optimized, ultra-low latency RAG and semantic search systems
- Improving data reliability through automated testing, validation, and CI/CD
- Crafting maintainable documentation and developer-friendly data architectures
