Skip to content
View SultanAlfaifi's full-sized avatar
13 followers · 6 following

Achievements

Achievement: YOLOAchievement: Pull Shark

Achievements

Achievement: YOLOAchievement: Pull Shark

Block or report SultanAlfaifi

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.

Content in all repositories owned by your account will be closed.
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
SultanAlfaifi/README.md

Sultan Alfaifi

AI Agent Builder · Applied AI Researcher

Building safe, explainable, and evaluation-driven AI systems for Arabic users.

Website · LinkedIn · X

Focus

  • Agent systems: tool use, structured outputs, memory, orchestration, and human approval.
  • Agent safety: bounded context, prompt-injection resistance, privacy, and review-before-action.
  • Applied AI research: reproducible experiments, evaluation, failure analysis, and honest limitations.
  • Arabic-first AI: useful systems designed around Arabic language and regional needs.

Featured agent systems

Project What it demonstrates
Mustakshif A privacy-first Windows app and CLI for explainable local-model recommendation using trusted source data and hardware-aware scoring.
AutoFlow An Arabic financial-automation agent prototype with text and voice interaction, structured workflow generation, safety constraints, and explicit approval before activation.
Sultan AI Skills Open, reusable Agent Skills for Codex and Claude with deterministic validation scripts.
DANNA An Arabic Alexa voice agent exploring contextual, multi-turn interaction with large language models.

Applied research

Project Research direction
MAWRID Safety-oriented greywater reuse research with prior-art positioning, measurable validation targets, fail-safe controls, and explicit limits on unvalidated claims.
KAUST AI Projects Computer-vision and deep-learning experiments covering classification, transfer learning, and semantic segmentation with PyTorch.

Research direction

My current direction is reliable Arabic AI agents: evaluating tool use, memory, approval boundaries, privacy, prompt-injection resistance, and failure behavior under realistic tasks.

I care about systems that can be inspected and tested—not just impressive demos. My public work aims to document architecture, methodology, security boundaries, reproducibility, and known limitations.

Background

  • IBM AI Developer Lab
  • McKinsey Forward Fellow
  • KAUST AI Specialization — Top 5%
  • Software Engineering, product design, and full-stack delivery
  • Based in Saudi Arabia

Working principles

  1. Measure behavior instead of relying on claims.
  2. Keep humans in control of consequential actions.
  3. Minimize and bound the context given to agents.
  4. Document failure modes and limitations.
  5. Build Arabic experiences as first-class products.

Open to research collaboration, open-source contributions, and applied AI-agent projects.

Pinned Loading

  1. mustakshif mustakshif Public

    Discover trusted local AI models that fit your Windows PC — bilingual desktop app and CLI.

    Python 10

  2. AutoFlow AutoFlow Public

    Arabic financial-automation agent prototype with voice, structured workflows, safety constraints, and approval-first execution.

    JavaScript 1

  3. ai-skills ai-skills Public

    Open, reusable Agent Skills for Codex and Claude with validation scripts.

    Python 1

  4. DANNA DANNA Public

    Arabic Alexa voice agent exploring contextual, multi-turn interaction with large language models.

    Python 1

  5. mawrid-water-reuse mawrid-water-reuse Public

    Smart decentralized reuse of ablution greywater using date-palm biochar, condition-based monitoring, and fail-safe control.

    HTML

  6. KAUST-AI-Projects KAUST-AI-Projects Public

    Built, trained, and optimized real-world AI models as part of the KAUST AI Specialization (Top 5%), focusing on Computer Vision, Deep Learning, and practical problem-solving using PyTorch.

    Python 2