I build full-stack systems, containerize them, deploy to the cloud, then train and secure what ships.
CS undergraduate at CUET, working across data science, machine learning, and AI engineering. I like taking a project from raw data to a deployed API, and I bring a security-first habit from a year as CISO at a Web3 agency.
Experience
Security leadership across community-driven Web3 products, alongside a CS degree in progress.
- Owned application security across several community-driven Web3 projects, including Harold Health, Undead Blocks, and LightningWorks. Ran threat models and vulnerability assessments before anything shipped.
- Worked with dev teams to build security into the SDLC instead of bolting it on at the end. Harold Health passed review, reached token launch, and is now listed on Dexscreener.
- Led a security and SEO audit of the agency's own web presence: found the problems, prioritized them, and followed through on fixing them.
- Translated technical risk findings into recommendations stakeholders and developers could actually act on.
Coursework and project work spanning data structures & algorithms, machine learning, databases, and software engineering, alongside the security and AI/ML work above.
Technical Skills
Data science and ML lead the way. Security work sits alongside it.
Languages
Data Science & ML
Visualization & BI
Backend & APIs
Databases & DevOps
Security
Projects
Filter by practice area, or browse everything below.
SIEM Log Pipeline + AI Anomaly Detection
Real-time log parsing and feature extraction feeding Isolation Forest and One-Class SVM models to flag anomalous activity. Sits right at the intersection of the ML and security work.
Cybersecurity Mini Projects
A small Python security toolkit: password auditor, port scanner, DNS lookup, JWT inspector, hash/hex tools, and a Windows event log parser.
Harold Health Security Audit
Identified and remediated a critical broken-access-control vulnerability in a patient-provider credential registry, patched with strict owner validation.
AI Resume Analyzer / BetteResume-ai
A REST API and full-stack app that parses resumes against a job description using NLP/LLM calls, scores ATS fit, and generates targeted feedback.
Customer Churn Prediction
Predicts customer churn using a Gradient Boosting model with SMOTE to handle class imbalance, served through a Dockerized Flask API kept separate from the model layer so either can change without touching the other.
Academic Performance Prediction
ML pipeline predicting student outcomes from behavioral and academic data using a Random Forest classifier.
Loan Approval Prediction
A classification pipeline for credit-risk assessment that compares SVM against a Decision Tree, aimed at automating what's normally a manual approval check.
RNN Name Classifier
Character-level RNN built from scratch in PyTorch with custom training vectors for name/nationality classification.
Automated Crypto Market Tracker
CoinGecko API integration tracking BTC, ETH, XRP, DOGE, and BNB with historical Excel storage and automated visual reporting.
COVID-19 Data Exploration
Global COVID-19 dataset analysis in Microsoft SQL Server: data cleaning, joins, and dashboard-ready aggregation.
Nashville Housing: Data Cleaning
Standardized inconsistent formats and removed duplicates in a real-world housing dataset using SQL Server.
Power BI: Data Professional Survey
Interactive dashboard profiling global data-professional salary benchmarks, language preferences, and demographics.
Movies Correlation Project
Explored relationships between budget, gross revenue, runtime, and ratings using heatmaps and scatter plots to find what drives revenue.
CredChain-Nexus
Blockchain-based academic credential verification system built for Bangladesh's education ecosystem.
PitchIt AI
AI sponsorship-outreach generator that pulls live research through Firecrawl and Tavily before drafting each pitch.
Certifications & Achievements
Certifications
- Google Cybersecurity Certificate
- ISC² Certified in Cybersecurity (CC)
- LetsDefend SOC Fundamentals
Achievements
- Top 938 Worldwide in picoCTF
- Former CISO at W3D, leading security across several Web3 projects
- Competed in IEEE Computer Society and Kaggle Datathon hackathons
Recommendations
A word from the people I've worked with directly.
Kareeb served as Chief Information Security Officer at Web3Discord.com, where he led application security initiatives, ran STRIDE threat models, and conducted vulnerability assessments across the agency's Web3 projects. He also supported data collection, web scraping, AI automation, KPI reporting, and website development, while handling Level 2 ticketing and technical support for the team. He worked well within the team, communicated clearly, and was dependable, which the letter cites as reason enough to recommend him for entry-level roles in data science, AI, or machine learning.
Specialized Portfolios
Deeper write-ups for the two data practices this site only summarizes, each with its own live site, project write-ups, and repo links.
Publications & Articles
Security write-ups and AI/ML articles, published on Medium.
W3D / Harold Health Security Report
Identified and remediated a critical broken-access-control vulnerability in a patient-provider credential registry.
Read on Medium →More AI/ML & Security Articles
Ongoing write-ups on machine learning projects, application security, and lessons from CTF competitions.
Visit Medium (@kareebsadab) →Let's work together
Open to internships, research opportunities, freelance projects, and full-time roles in Cybersecurity, AI/ML, Data Science, and Backend Engineering.