Maria Habib

Maria Habib

Full-Stack Engineer & Research Specialist in Program Analysis

Based in: Lahore, PakistanMost recently: Teaching Assistant - Algorithms, Statistics & Networks, PUCIT, University of the PunjabExperience: 1 year

Hire me

Information Technology graduate from PUCIT with extensive experience in full-stack development and research engineering using Python, React, and Node.js. Proven track record in building complex code-similarity platforms and deterministic evaluation frameworks for AI/ML products.

Experience

Teaching Assistant - Algorithms, Statistics & Networks

PUCIT, University of the Punjabpart time2025 – Jun 2026
Evaluated 100+ student submissions for correctness, logic, and specification adherence while managing concurrent assessment cycles.
AlgorithmsStatisticsNetworks

Software Engineer Intern

Spacebar TechnologiesinternshipSep 2025 – Nov 2025

Summary

Contributed to full-stack development using React.js and Git/GitHub in an Agile/Scrum team through sprint work, code review, and ticket-driven delivery.

What I did

  • Connected the frontend with the authentication flow and handled cases like invalid or unauthorized requests.
  • Tested the authentication integration and resolved issues as they arose.

Results

  • Ensured that only authenticated users could access the messaging functionality by implementing message authentication for a chat application.
React.jsGitGitHubAgileScrum

Digital Twin Intern - IoT Digital Twin Project

Research ProjectinternshipJun 2025 – Aug 2025

Summary

Validated sensor-data streams and backend integration, reconciled real-time outputs, investigated ingestion edge cases, documented defects, and verified fixes. Contributed to reverse engineering of Eclipse Ditto Java components, identifying and cataloguing code smells, bugs, and quality violations with documented audit trails.

What I did

  • Reverse engineered the Eclipse Ditto open-source codebase to explore the Java source code.
IoTEclipse DittoJavaSensor-data streamsSonarQubeCodacyGit

Skills

Technical

Source-code Similarity
Git/GitHubGit/GitHub
PythonPython
REST APIs
HTML5HTML5
CSS3CSS3
SQLSQL
Node.js/ExpressNode.js/Express
Experimental Evaluation
FastAPIFastAPI
AST/Token Representations
Agile/Scrum
PostgreSQLPostgreSQL
React.jsReact.js
JavaScript (ES6+)JavaScript (ES6+)
Tree-sitter
ViteVite
MySQLMySQL
SupabaseSupabase
Schema Design
Microservices
LinuxLinux
JiraJira
Codacy
CI/CD
DjangoDjango
SonarQube
token-normalization
structure-aware matching
Static Analysis
Independent project creation
Reverse Engineering
Rapid prototyping
Independent Research

Projects

CodeSpectra - Academic Code Similarity Platform (FYP)

Full-Stack Developer / Architectacademic
Designed and built a multi-stage full-stack platform combining exact/token-normalized comparison, AST/LCS-based structural analysis, and behavioral analysis. Architected role-based workflows, REST APIs, PostgreSQL-backed data flows, side-by-side code comparison, and automated assessment features across React, Node.js, and Python services. Reported processing of 100 submissions in about 4.2 minutes in project evaluation; the project was featured on the Ptidej Research Blog.
ReactViteNode.jsExpressFastAPIPostgreSQLPythonASTLCS

SimProbe - Structural Robustness in Source-Code Similarity

Research Engineeracademic
Built a deterministic evaluation framework comparing token-normalized and method-aware source-code similarity under controlled transformations and external benchmark data. Ran a frozen 6,000-pair BigCloneBench evaluation (5,755 measurable pairs) and maintained reproducible protocols, deterministic sampling, experiment artifacts, and correction provenance. Identified a clear evaluation boundary: method-aware matching improves robustness to method reordering and class splitting, but provides little advantage when inputs contain only one method.
PythonProgram AnalysisBigCloneBenchtoken-normalizationstructure-aware matching

AI-Driven Intrusion Detection System (Mini-IDS)

Developerpersonal
Developed a real-time monitoring prototype handling 700+ events/second with under 100 ms detection latency, including alert triggering, confidence tracking, and incident logging.
PythonJavaScriptREST API

AI-driven honeypots

personal
Independent research paper on AI-driven honeypots supervised by a university professor as an information security project.
ResearchGate

BehavClone

personal
A project that explores behavioral code similarity, with the implementation and experiments available publicly on GitHub.
behavioral code similarity

Education

B.S. Information Technology

University of the Punjab (PUCIT)Information TechnologyGPA 3.522022 – 2026

Coursework

AlgorithmsStatisticsNetworks

Recognition

Certifications

Python Programming

DataCamp

Introduction to SQL

DataCamp

Intermediate SQL

DataCamp

Patents & publications

CodeSpectra: The Illusion of Easy Coding - Why AI Still Demands Effort

publication2026

Research publication regarding AI and code similarity.

With Habib, M., Razzaq, M.

AI-Driven Honeypots and Deception Technologies in Modern Cybersecurity

publication2026

Research publication on cybersecurity deception technologies.

With Habib, M.

Hire MariaReach out about a role, a contract or a conversation.For recruiters

Maria's twin is AI, it can make mistakes.

More people on Kanvis

Browse more peopleCreate your own Kanvis