Welcome!
Islam

Hey, I'm Islam!

A Duke University student interested in AI/ML pipelines and infrastructure. Currently based in Durham, NC.

I grew up between Egypt and Saudi Arabia. I'm a big fan of Alternative Hip Hop and enjoy playing Tetris and Monkeytype in my free time.

Experience

ML Research Assistant

Incoming

Duke UniversityML + infrastructure for enzyme design @ Romero Lab

Durham, NC
Responsibilities & Results:
    PythonPyTorchMachine LearningContinual LearningDistributed TrainingDomain GeneralizationHPCBioinformatics

    Software Engineer Intern

    Present
    Backed by YC

    SoffAI + infrastructure for supply chain intelligence

    May 2025 – Present
    San Francisco, CA
    Responsibilities & Results:

      Software Engineer

      HelianDeveloping AI insight tool for medical research workflows

      Dec 2024 – May 2025
      Durham, NC
      Responsibilities & Results:
      • Engineered ETL pipeline achieving 80% speedup and 65% resource reduction
      • Built document processor embedding 25K daily PDFs with <100ms latency
      • Implemented caching system reducing costs by 15% and storage by 10%
      • Optimized RAG reducing token usage by 25% and response time by 40%
      TypeScriptPythonNext.jsFastAPISupabaseAWSRedisCeleryDockerGit

      ML Engineer Intern

      Reveal GenomicsBreast cancer genomic analysis & biomarker discovery

      Sep 2024 – Apr 2025
      Durham, NC
      Responsibilities & Results:
      • Enhanced biomarker identification by 55% with Dask + NetworkX pipelines
      • Improved non-linear gene correlation detection by 65% using PCA + t-SNE
      • Reduced R&D analysis time by 85% with custom genomic dashboards
      PythonPandasStreamlitDaskNetworkXScikit-LearnPCARandom ForestTF-IDFt-SNEGit

      ML Research Assistant

      Duke UniversityML for therapeutic protein design @ NaderiAlizadeh Lab

      Oct 2024 – Apr 2025
      Durham, NC
      Responsibilities & Results:
      • Engineered continual learning model increasing generalization by 9%
      • Built distributed pipeline optimizing 1M+ protein candidates with 20% speedup
      • Implemented GearNet integration boosting prediction accuracy by 20%
      PythonPyTorchMachine LearningContinual LearningDistributed TrainingDomain GeneralizationHPCBioinformatics

      Software Engineer Intern

      Duke Institute for Health InnovationAutomated health literature review system

      Jun 2024 – Aug 2024
      Durham, NC
      Responsibilities & Results:
      • Engineered review system processing 250+ papers daily
      • Built classifier achieving 98% accuracy and 95% faster processing
      • Implemented grant-writing assistant serving 350+ analysts across organizations
      • Developed BERT detector analyzing 10K+ EHR records hourly
      PythonBERTAutoGenLLMsGROBIDEHR AnalysisDocker

      Software Engineer Intern

      Project: SapienNLP + population health analysis dashboard

      Dec 2023 – Jan 2024
      Princeton, NJ
      Responsibilities & Results:
      • Engineered survey builder reducing creation time by 25%
      • Built classification pipeline reducing analysis time by 95%+
      • Implemented HIPAA-compliant data anonymization ensuring privacy
      JavaScriptReactNode.jsBERTRegexData PrivacyHIPAA

      ML Research Assistant

      Saudi Aramco, KFUPMMetal-organic polymers for CO₂ capture @ IRC-HTCM

      Jul 2022 – Sep 2023
      Saudi Arabia
      Responsibilities & Results:
      • Engineered simulations predicting capture capacity within 12% of results
      • Built meta-analysis analyzing 300+ papers for proposals (3 publications)
      • Implemented breakthrough analysis interface processing 36K+ points hourly
      MATLABPythonStreamlitScikit-LearnSeabornData VisualizationMonte CarloMeta-Analysis
      Selected Projects

      In Progress
      Better Finder

      Developing a native macOS Finder alternative for developers with AI-powered search, customizable workflows, and advanced filtering capabilities. Designed to offer power users with maximal automation and productivity.

      • Swift
      • SwiftUI
      • Core ML
      • macOS API
      • Metal
      • FileProvider
      Better Finder

      Etchr – GitHub README Generator
      New!

      Web application that reduces README creation time from 120+ minutes to 5 clicks, serving 100+ users with 65% repeat usage by leveraging Google's Gemini AI to analyze codebases and generate documentation.

      • TypeScript
      • Next.js
      • Express.js
      • Node.js
      • Supabase
      • Google Gemini
      • GCP

      Job Track – CLI Job Tracker

      CLI tool that automates the tracking of job applications by extracting information from Gmail emails using Google Gemini AI and updating a Google Sheet, eliminating manual data entry.

      • Python
      • Google Gmail API
      • Google Sheets API
      • Google Gemini
      • CLI
      See more on GitHub...
      Technical Skills
      CategoryTechnologies
      Programming Languages
      Machine Learning
      Frontend Development
      Backend Development
      Deployment & Cloud
      Selected Publications

      Dec. 2024

      Literature Review
      Hamid Zentou, Mansur Aliyu, Mahmoud A. Abdalla, Omar Y. Abdelaziz, Bosirul Hoque, Ahmed M. Alloush, Islam M. Tayeb, Kumar Patchigolla, Mahmoud M. Abdelnaby
      Carbon CaptureClimate MitigationMaterials Science

      Dec. 2023

      Research Article
      Mahmoud Abdelnaby, Islam Tayeb, Ahmed Alloush, Hussain Alyosef, Aljazi Alnoaimi, Mostafa Zeama, Mohammed Mohammed, Sagheer Onaizi
      Organic ChemistryMaterials ScienceEnvironmental Tech

      Jul. 2022

      Case Report
      Abdullah Alsulaiman, Siraj Alharthi, Ahmed Albariqi, Rasha Mutabaqani, Fawzi Bokhari, Islam Tayeb, Dalia Alharthi, Muhammad Tariq, Yasser Babaier
      Medical OncologyClinical GeneticsRegional Studies

      See more on Scholar...
      Contact
      Email
      LinkedIn
      GitHub
      X
      Google Scholar
      Resume