 |
 |
 |
 |
const ABDEL_ATIA = {
education: {
degrees: ["PhD (Biopharmaceutics)", "DMV (Veterinary Medicine)"],
certifications: [
"IBM RAG & Agentic AI — 8-course specialization",
"Harvard CS50 — 7 courses completed"
]
},
coding: {
started: "Childhood — self-taught since day one",
evolution: "BASIC → C → Python → AI/ML → RAG & Agents",
experience: "15+ years"
},
stats: {
countries_visited: 64,
continents_explored: 4,
certifications_earned: 8,
papers_read_yearly: 300,
coffee_dependency: "CRITICAL ☕"
},
superpower: "I speak the language of cells, Python, AND LLMs.",
motto: "Automate the boring. Amplify the human."
} as const; |
Certified by 
.png)
|
🔬 My Journey — From childhood coder to AI architect
+ 👦 Fell in love with code as a child
+ 🖥️ Self-taught out of pure curiosity
+ 📚 From BASIC to modern languages |
+ 🦁 DMV — Treated lions in Africa
+ 🧬 Biopharmacist — Molecular research
+ 📜 PhD — Biopharmaceutical sciences |
+ 💻 Fusing Biology + Code
+ 🤖 AI Systems Engineer
+ 🎯 IBM + Harvard (2025-2026) |
The Common Thread: Code has been with me since childhood. Biology gave me scientific rigor. AI lets me fuse it all together.
My OMML→MathML converters (radical/sqrt, group char, phantom, n-ary, eqArr, accents, pre-sub, limits) are in the v1.27.0 stable release notes.
Snapshot: April 2026 · Totals from
author:Abdeltoto is:pr is:merged · 14 PRs merged into third-party repos + 2 on my own public repos (below)
|
superdoc-dev/superdoc
 Built the full OMML → MathML conversion layer: equation arrays, matrices, n-ary operators, group characters, phantoms, accents, pre-sub-superscripts, and box/borderBox containers. Enables Word math equations to render natively in the browser. Featured in the v1.27.0 stable release notes (2026-04-17). #2754 · #2753 · #2752 · #2751 · #2750 · #2749 · #2748 · #2747 |
|
palantir/palantir-mcp Fixed cross-platform compatibility — Palantir's MCP server couldn't spawn on Windows. Added shell: true to the npx child process, unblocking all Windows developers.#25 |
|
DeepCritical/DeepCritical Implemented a full MCP server for MAFFT multiple sequence alignment — bridging bioinformatics tooling with the AI agent ecosystem. Allows LLM agents to run protein/DNA alignments autonomously. #231 |
|
microDM/ProBioPred Modernized the project scaffolding — added pyproject.toml, CONTRIBUTING.md, and .gitignore to make this probiotic prediction tool pip-installable and contributor-friendly.#1 |
|
Awesome Bioinformatics & Computational Biology Enriched two major curated lists used by thousands of researchers — added Scanpy (the standard scRNA-seq toolkit) and EvoDiff (protein generation via diffusion models). Awesome-Bioinformatics #127 · awesome-computational-biology #50 |
|
rmindel/gsd-for-cursor Fixed broken hooks path — the Cursor IDE adaptation was still pointing to ~/.claude/ instead of ~/.cursor/, causing silent failures for all Cursor users.#3 |
|
Abdeltoto/vllm-omni_AA #1 — [CI/Build] Add L5 stability and reliability test framework |
|
Abdeltoto/FacilityMapper_dbg #1 — Add .gitignore
|
RAG Systems Documents → Knowledge → Action Hybrid search, re-ranking, citations |
AI Agents Multi-agent orchestration ReAct, MCP, autonomous workflows |
Bioinformatics FASTQ → Species identification 16S rRNA, BLAST, metagenomics |
Production APIs FastAPI, Docker, K8s Scalable, monitored, responsible |
8-Course Specialization | Advanced Level 
|
7 Courses Completed | Full-Stack Computer Science Curriculum 
|
timeline
title 🎯 Certification Journey
2025 : 🎓 CS50x — Computer Science
: 🐍 CS50P — Python Programming
2026 : 🗄️ CS50 SQL — Databases
: 🤖 CS50AI — Artificial Intelligence
: 📊 CS50R — R Programming
: 🌐 CS50W — Web Programming
: 🔒 CS50 Cybersecurity
: 🤖 IBM RAG & Agentic AI — 8 Courses
: 🚀 More coming soon...
-00D9FF?style=for-the-badge){kind=icon}
Stack:
- LangChain + LangGraph
- Pinecone + Cohere Rerank
- GPT-4.1
Features:
✅ Multi-source ingestion
✅ Hybrid search (dense + sparse)
✅ Self-correcting RAG
✅ Citation & provenance
Impact: "90% faster research" |
Stack:
- LangGraph + CrewAI
- Claude 4 + GPT-4.1
- Custom Tools + MCP
Agents:
✅ Researcher (literature)
✅ Analyst (data)
✅ Writer (synthesis)
✅ Critic (review)
Impact: "20h/week automated" |
Stack:
- Python + Flask (SPA)
- NCBI BLAST+ & SILVA 138.2
- Chart.js + fpdf2 reports
- SQLite + Auth multi-users
Features:
✅ 16S rRNA metabarcoding pipeline
✅ Paired-end R1+R2 assembly (Phred)
✅ PERMANOVA, ANOSIM, PCoA, Bray-Curtis
✅ Rarefaction curves & taxonomic trees
✅ PDF reports + BIOM 1.0 export
✅ SaaS: Free / Pro / Admin plans
Impact: "FASTQ → Species ID in minutes" |
Stack:
- Airflow + Docker
- PostgreSQL + Redis
- Custom monitoring
Features:
✅ Self-healing workflows
✅ Anomaly detection
✅ Auto-scaling
✅ Real-time alerts
Impact: "Zero manual intervention" |
class DevPhilosophy:
RULES = [
"If you do it more than twice → automate it",
"Clarity beats cleverness. Always.",
"Ship fast, iterate faster, learn fastest",
"Good code is code that can be deleted",
"The best prompt is the one you never write twice"
]
MOTTO = "In the name of the Prompt, the Model, and the Holy Iteration 🙏"
def daily_standup(self) -> str:
return """
☕ Coffee.init()
💻 Code.write()
🐛 Bugs.hunt()
🚀 Features.ship()
📚 Papers.read()
🔄 Repeat()
""" |
- ⚠️ CaffeineDependency: System unstable below 3 cups
- ⚠️ TabOverflow: 50+ browser tabs is "normal"
- ⚠️ RefactorLoop: "Quick fix" → 5h rewrite
- ⚠️ CertificationAddiction: 8 and still collecting
+ ✅ RAGObsession: Building retrieval systems 24/7
+ (working as intended)
+ ✅ AgentAddiction: Creating AI agents for everything
+ (feature, not bug)
+ ✅ HarvardEffect: Now explains everything with O(n)
+ (feature)
+ ✅ CS50Completionist: All 7 CS50 courses done
+ (definitely not a bug) |
