Hi, my name is

Abdalla Mohamed.I build autonomous AI agents.

Full-Cycle AI Technical Product Engineer. I design, evaluate, and productionize autonomous agents — from research and prototyping all the way to production and real customers. I care less about demos and more about evaluation science: the reliability work that makes autonomous systems trustworthy at scale.

Now
Senior AI Engineer
@ DataCamp · optimizing a top-rated edtech AI tutor
Founder & CEO
Bareq
Agentic AI solutions agency · 5 clients shipped
Founding R&D
Stakpak
Open-source DevOps coding agent · Silicon Valley
scroll ↓
01. About me

Abdalla is a Full-Cycle AI Technical Product Engineer with 4+ years across software engineering, AI research, and technical education. He is currently a Senior AI Engineer at DataCamp, optimizing one of edtech's top-rated AI tutors through evaluation science and A/B testing.

As a former founding product engineer at Stakpak for a year, he operated in the heart of San Francisco's Silicon Valley ecosystem, competing head-to-head in R&D and product engineering against frontier agentic tools like Claude Code and OpenCode. He specializes in designing and optimizing agentic tools — from real-time agentic search and context engineering to coding-agent evaluation frameworks, hallucination detection, auto-prompt optimization, and intelligent retrieval systems that make AI agents more reliable and capable. Along the way he's built deep hands-on exposure to DevOps and cybersecurity — from Kubernetes and multi-platform deployment to security-hardened, self-hosted agent systems.

With 4+ years of combined experience across production AI systems, software engineering, and technical education — including a Senior Teaching Assistant role at the German University in Cairo — he has shipped and supervised work spanning ML, embedded systems, and autonomous systems. Previously the founding AI Engineer at Seqoon, he led a team shipping production RAG chatbots and a real-time Egyptian-Arabic voice assistant for real estate, and built an Islamic-books Notebook LLM (hybrid search, multi-agent summarization, podcast generation) as a freelance project.

A Mechatronics Engineering graduate (Excellent with High Honors) with an A+ thesis on Deep Reinforcement Learning for autonomous-vehicle navigation and an exchange semester at GUC Berlin. Beyond building products, he's passionate about growing AI communities — organizing hackathons, giving talks, and creating spaces where engineers learn and build together. He believes the best AI products come from engineers who care equally about the research, the craft, and the people using them.

A few things I work with often:

  • Real-time agentic search
  • Agent evaluation & LLM-as-judge
  • Context engineering
  • Multi-agent orchestration
  • Hallucination detection
  • Fine-tuning (LoRA/QLoRA)
  • RAG with fresh retrieval
  • Agent recovery & reliability

At a glance

📍 Cairo, Egypt · remote-friendly

🎓 B.Sc. Mechatronics, GUC — 3.85/4.0, High Honors

🗣️ English · Arabic (native) · German (A2)

🎤 Lead instructor — AI Engineering Workshop Series

🪂 Off-screen: military parachuting & skydiving, squad leader

02. Where I've worked

Experience

A detailed look at each role — what I built, and the impact it had.

Senior AI Engineer · DataCamp
May 2026 — Present
Remote
  • Optimize DataCamp's AI-native course tutor — among the most efficient, top-rated AI tutors in edtech — via sustained evaluation cycles: curating datasets that close blind spots and A/B testing across models and system designs for measurable quality gains.
  • Shipped the platform guidance feature A→Z into production — full-stack from a Python/FastAPI backend to an Alpine.js + HTMX interface — plus escalation detection and reporting for reactive AI behaviors.
  • Built analytics & evaluation pipelines on GCP BigQuery to mine learner-interaction data, quantify tutor performance, and prioritize AI product improvements.
EvaluationA/B testingFastAPIHTMXBigQuery
Founding R&D AI Software Engineer · Stakpak (acquired by Vercel)
Mar 2025 — Mar 2026
San Francisco, CA (Remote) · Founding team · Open-source agent
  • Full-cycle AI product engineer: identified product gaps, led research, implemented in Rust & Python, and onboarded customers. Designed core agentic components — tool-calling interfaces, context engineering, dynamic memory, caching, and multi-agent orchestration for a 24/7 autonomous infrastructure agent.
  • Architected an end-to-end agentic search & retrieval system (Search Docs) in Rust — caching, semantic reranking, trusted-domain filtering, iterative LLM refinement — hitting +40% recall and +55% precision with guaranteed freshness, after benchmarking Crawl4AI, Playwright, and Context7 MCP.
  • Pioneered Stakpak's first DevOps evaluation dataset (Harbor / Terminal-Bench) with containerized tasks and automated verification; built an automated agent-testing framework with LLM-as-judge (Langfuse) and trajectory analysis over 1000+ sessions.
  • Designed a message versioning & checkpointing system for the agent — with automatic recovery detection that, on detecting a violation, rolls back to the last safe checkpoint and switches the agent automatically, keeping long-running autonomous sessions reliable.
  • Built an automatic prompt optimizer on top of the eval harness — extracting the loss signal, mapping it to the responsible prompt section, and rewriting it in a loop to preserve performance while optimizing toward the global maximum. Also designed a skill / rulebook generator that analyzes production sessions to codify DevOps playbooks (Kubernetes, Cloudflare, Netlify, Coolify, vLLM deployment).
  • Designed auto-prompt-optimization (DSPy/Adalflow-inspired) and hallucination detection/recovery; ran Qwen 3B fine-tuning (Unsloth, LoRA/QLoRA); co-developed the 24/7 Autopilot mode.
RustAgentic retrievalEvalsMulti-agentLoRA/QLoRA
Founding AI Software Engineer · Seqoon (hiresai.ai)
Nov 2024 — Jun 2025
Cairo, Egypt
  • Owned and shipped two production AI products end-to-end alongside 3 engineers. Architected SAI Dubai — a bilingual (AR/EN) real-estate chatbot on a full RAG pipeline (Milvus, Pinecone, Neo4j) with multi-turn LangChain/LangGraph agents, chunking, embeddings, vector indexing, and metadata filtering (FastAPI, MongoDB).
  • Prototyped SAI Sales Agent — an Egyptian-Arabic voice AI: Twilio telephony + LiveKit, TTS/STT pipelines, dialect fine-tuning (Whisper, XTTS, Sesame), and low-latency local inference via vLLM & Ollama.
RAGLangGraphVoice AIvLLM
Senior Teaching Assistant · German University in Cairo
Sep 2022 — Nov 2024
Computer Science & Engineering Dept · Cairo, Egypt
  • Taught C/C++/Python, Computer Architecture, Embedded Systems, and Autonomous Systems; led 500+ student projects and supervised 10+ Bachelor theses in deep learning for automotive (CNNs, Transformers, LSTMs, DRL).
TeachingDeep learningEmbedded
03. Things I've built

Selected projects

Engineering work across retrieval, evaluation, and production agent systems.

Agentic Search & Retrieval — “Search Docs”

Stakpak · founding R&D

A live search & retrieval system in Rust feeding a 24/7 autonomous DevOps coding agent — real-time web search backed by a caching layer for speed and freshness, with semantic reranking, trusted-domain filtering, and iterative LLM-based refinement to keep only the most relevant, current context in front of the agent. Benchmarked crawling & retrieval strategies (Crawl4AI, Playwright, Context7 MCP) against defined speed, freshness, recall, and precision metrics — landing at +40% recall and +55% precision with guaranteed data freshness.

RustLive web searchCachingSemantic rerankingLLM refinementEvaluation

Agent Recovery & Checkpointing

Stakpak · founding R&D

A message versioning & checkpointing system for the agent that makes long-running autonomous execution reliable. I designed the checkpointing scheme over the agent's message history plus an automatic recovery-detection layer: when a violation is detected, the system rolls back automatically to the last safe checkpoint and switches the agent to recover — no human babysitting required for the agent to get itself back on track.

RustCheckpointingMessage versioningAutomatic rollbackReliability

Automatic Prompt Optimizer

Stakpak · founding R&D

An auto-prompt-optimization layer built on top of the evaluation harness: it extracts the loss signal from evals, maps it back to the responsible section of the prompt, and then automatically rewrites that section in a loop — preserving current agent performance while pushing toward the global performance maximum. Inspired by DSPy and Adalflow, with textual-gradient refinement and multi-tier evaluation.

Prompt optimizationTextual gradientsEval-drivenDSPy / Adalflow

Skill / Rulebook Generator

Stakpak · founding R&D

A system that analyzes production agent sessions and auto-designs reusable skills and rulebooks for the agent — codifying DevOps standard operating procedures for Kubernetes deployment, Cloudflare, Netlify, Coolify, and vLLM deployment so the agent applies proven, repeatable playbooks instead of improvising.

RulebooksKubernetesCloudflare / Netlify / CoolifyvLLM
🧪

Agent Evaluation Framework

Stakpak

Automated agent-testing with full A/B cycles: dataset generation, LLM-as-judge pipelines with Langfuse observability, and sub-goal trajectory analysis over 1000+ coding sessions — surfacing failure patterns that end-to-end testing missed. Includes Stakpak's first DevOps eval dataset on the Harbor / Terminal-Bench standard.

LLM-as-judgeLangfuseHarborTerminal-Bench
🏙️

SAI Dubai & SAI Sales Agent

Seqoon · founding engineer

A bilingual (AR/EN) real-estate chatbot on a full RAG pipeline (Milvus, Pinecone, Neo4j) with multi-turn LangChain/LangGraph agents; plus an Egyptian-Arabic voice AI for property sales — Twilio + LiveKit, TTS/STT, dialect fine-tuning (Whisper, XTTS, Sesame), low-latency inference via vLLM & Ollama.

RAGLangGraphVoice AIvLLM
04. My company

Bareq

The agentic AI solutions agency I founded and lead — from architecture to deployment.

Bareq
Bareq · بارق
Founder & CEO · 2025 – Present · bareq.dev

An agentic AI software agency delivering production-grade autonomous systems. I own delivery end-to-end — architecture, build, and deployment — and have closed and shipped for 5 clients to date. We build the governance layer around AI, not just the demo: self-hosted models, evaluation, security, and human-in-the-loop by design.

SOC / cybersecurity agentsHR-department automationR&D process automationIntelligent CRMsAI-native web apps
Bareq — selected products

SOC AI Analyst Co-Pilot

In production

Self-hosted agentic co-pilot wired into the SIEM, incident DB, and threat-intel stack; every verdict grounded in live tool calls against the client's own data. Grounded triage verdicts with evidence in under 30s (~90% faster time-to-triage) via parallel investigation — role-aware tiers (Tier 0→2, MITRE ATT&CK), strict memory isolation, redacted audit trail, human-in-the-loop.

Self-hosted LLMsSIEMMulti-agentSecurity

Sarmad — AI Islamic Library

Lead engineer

A “NotebookLM for hadith & Islamic literature”: customizable notebook mode, full-database hybrid search, and voice-enabled reading. OCR & scraping pipelines for digitizing books; hybrid retrieval (Typesense BM25 + embeddings + Gemini long-context); a multi-agent system with agentic RAG, summarization, and podcast-script generation. Led a team of 2–3 engineers.

ReactDjangoSupabaseHybrid RAG
05. Events, speaking & community

Talks & community

Jun – Jul 2026
AI Engineering Workshop Series · Lead instructor & organizer
SOIL Coworking, Cairo · 4 sessions · 16+ teaching hours. Designed & led the full curriculum — from GPT foundations to shipping production agents — with a team of industry mentors, closing with a 20–25-team hackathon night.
Playlist ↗
2025
YouTube Live with Ahmed El Emam · Guest speaker
Live discussion on AI engineering, coding agents, and career paths in agentic AI.
Watch ↗
2025
DevOpsDays Cairo 2025 · Technical speaker
Live product demo & workshop: the Stakpak agent dockerizing a web application in under 15 minutes.
Talk
2025
GenAI Hackathon Cairo 2025 · Co-organizer & workshop leader
Co-organized a 150+ participant hackathon; delivered a workshop on LLM workflows and coding agents; recruited judges from Vectara, Valeo, and Thundr.
Organizer
AI Engineering Workshop Series — session breakdown
01 · FOUNDATIONS
From Tokens to GPT
20 Jun 2026 · ✓ delivered
  • ML foundations — perceptron → RNNs/LSTMs, tokens & NLP basics, and the transformer mindset.
  • Intro to AI system design; how LLMs actually work for developers.
02 · BUILDING AGENTS
Invading the Agent's Den
04 Jul 2026 · ✓ delivered
  • Transformer deep-dive + building a GPT from scratch.
  • Context engineering & harness — RAG, MCP, DB knowledge; agent architecture patterns; designing agentic systems.
03 · AGENTS IN PRODUCTION
Scaling, Eval & Observability
11 Jul 2026
  • Computer-use agents; agents & infrastructure opsDevOps, sandboxing, safe execution.
  • Prompting, latency, caching & tokens; evaluation & observability; AI product & growth for engineers.
04 · HACKATHON NIGHT
Pitches, Judging & Awards
18 Jul 2026
  • 20–25 teams pitch (5 min each), judging & deliberation.
  • Closing mini-talks from startup leaders, then the award ceremony.
06. Tools of the trade

Technical stack

Languages

Python · Rust · TypeScript · JavaScript · Modern C++/C · Java · SQL · Bash · VHDL

LLM & Agents

LangChain · LangGraph · Langfuse · DSPy · Adalflow · vLLM · Ollama · Unsloth · MCP · LoRA/QLoRA · LLM-as-judge

Retrieval & Search

Milvus · Pinecone · Neo4j · Typesense · Hybrid BM25 + embeddings · Semantic reranking · RAG

ML / DL

PyTorch · TensorFlow · Keras · Scikit-Learn · Pandas · NumPy · NLTK

Web & Platform

FastAPI · Axum · Django · React · Alpine.js · HTMX · Supabase · PostgreSQL · MongoDB · GCP BigQuery

DevOps & Infra

Docker · Kubernetes · K3s · Harbor · Traefik · NGINX · Cloudflare · Netlify · Coolify · vLLM deployment · AWS · GCP · CI/CD · Prometheus