Skip to content
View abdalla1912mohamed's full-sized avatar
19 followers · 7 following

Achievements

Achievement: StarstruckAchievement: Pull Sharkx2Achievement: YOLOAchievement: Quickdraw

Achievements

Achievement: StarstruckAchievement: Pull Sharkx2Achievement: YOLOAchievement: Quickdraw

Block or report abdalla1912mohamed

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
abdalla1912mohamed/README.md

Hi there 👋

Abdalla Mohamed

Agentic AI Engineer | Autonomous Systems | AI Infrastructure

I design, evaluate, and productionize autonomous coding agents.

Currently building open-source DevOps AI agents at Stakpak, focusing on long-running autonomous systems, evaluation science, and reliable agent execution.

🧠 What I Specialize In

  • Autonomous coding agents (24/7 infra agents)
  • Agent evaluation frameworks & benchmarking datasets
  • Context engineering & dynamic memory systems
  • Multi-agent orchestration
  • Auto-prompt optimization systems
  • Hallucination detection & recovery pipelines
  • RAG systems with trusted, fresh retrieval
  • Fine-tuning small LLMs (LoRA / QLoRA)
  • Long-session reasoning reliability

🚀 Current Work — Stakpak

Founding AI R&D Engineer building a DevOps coding agent in Rust competing in the autonomous CLI agent space.

Core Contributions

  • Architected agentic search & retrieval system (trusted domains, semantic reranking, freshness guarantees)
  • Designed and built internal DevOps evaluation datasets (Terminal-style containerized benchmarks)
  • Implemented automated A/B testing pipelines with LLM-as-judge scoring
  • Built sub-goal trajectory analysis to detect reasoning failure patterns
  • Designed hallucination detection supervisors for real-time agent correction
  • Developed auto-prompt optimization framework (textual gradient refinement + multi-tier evaluation)
  • Contributed to Autopilot mode (long-running autonomous infrastructure agent)

🎯 Current Research Interests

  • Reliable autonomous coding agents
  • Evaluation science for LLM systems
  • Context & tool design engineering
  • Prompt optimization algorithms
  • Agent observability & telemetry
  • Infrastructure-aware AI systems

🏗 Previous Experience

Seqoon — Built and deployed production RAG and voice-agent systems as Founding AI Engineer.

German University in Cairo (GUC) — Senior Teaching Assistant in CS & Mechatronics, supervising deep learning and autonomous systems projects.

🛠 Technical Stack

Languages

Python • Rust • C/C++ • SQL • Bash

LLM & Agents

LangChain • LangGraph • DSPy • Adalflow
vLLM • Ollama • Unsloth
LoRA / QLoRA fine-tuning
Multi-agent systems

Retrieval & RAG

Milvus • Pinecone • Neo4j • Typesense
Hybrid search (BM25 + embeddings)
Semantic reranking
Metadata-aware retrieval

DevOps & Infra

Docker • Kubernetes • K3s
Harbor • Traefik • NGINX
CI/CD • AWS
Prometheus • Trivy • Checkov

Backend & Tooling

FastAPI • Django • Axum
MongoDB • PostgreSQL
Streamlit • Git • Linux

🌍 Links

LinkedIn: https://linkedin.com/in/abdalla1912mohamed
Stakpak Agent: https://github.com/stakpak/agent

🚀 Languages & Tools

Python C++ C Embedded C SQL JavaScript VHDL Java Rust

🚀 Technologies and Frameworks

LangChain Hugging Face PyTorch TensorFlow React MATLAB Git SolidWorks CasADi AUTOSAR Linux PostgreSQL Django

Pinned Loading

  1. stakpak/agent stakpak/agent Public

    Ship your code, on autopilot. An open source agent that lives on your machines 24/7 and keeps your apps running. 🦀

    Rust 1.4k 135

  2. stakpak/genai-agent-hackathon-cairo-2025 stakpak/genai-agent-hackathon-cairo-2025 Public

    Two technical workshops from GenAI Hackathon Cairo 2025: build LLM apps with RAG & tools , and create a coding AI agent from scratch (Python).

    Jupyter Notebook 43 11

  3. Deep_reinforcement_learning_Bachelor_project Deep_reinforcement_learning_Bachelor_project Public

    Deep Reinforcement Learning navigation of autonomous vehicles. Implementation of deep-Q learning, dyna-Q learning, Q-learning agents including SSMR(Skid-steering_mobile robot) Kinematics in various…

    Python 12 4

  4. Block-Worlds-Problem-Solved-Using-A-Algorithm-on-ROS2-Platform Block-Worlds-Problem-Solved-Using-A-Algorithm-on-ROS2-Platform Public

    The Block World Problem is a foundational problem in artificial intelligence (AI) and robotics that serves as a framework for exploring concepts related to planning, reasoning, and problem-solving.

    CMake

  5. Attention_is_all_you_need_reimplementation Attention_is_all_you_need_reimplementation Public

    Implementing the infamous paper " attention is all you need " . This repository contains a from-scratch PyTorch implementation of the Transformer model as introduced in the seminal paper "Attention…

    Python 1

  6. Vehicle-Telemetry-Simulation-in-C- Vehicle-Telemetry-Simulation-in-C- Public

    This repository contains a simulated car instance built from various classes representing sensors, ECUs (Electronic Control Units), a Diagnostic ECU, and a central Logger. The system's architecture…

    C++