Syeda Zahra Ali Tirmizey

Distinguished Computer Science Student and AI Research Intern

Based in: Lahore, PakistanMost recently: Summer Research Intern, CISPA Helmholtz Center for Information SecurityExperience: 2 years

Hire me

A high-achieving Computer Science student at LUMS with a 3.73 CGPA and extensive experience in software engineering and AI research. She has successfully developed end-to-end internal applications using FastAPI and Azure, and conducted advanced research on reconstruction attacks at the CISPA Helmholtz Center.

Experience

Summer Research Intern

CISPA Helmholtz Center for Information SecuritySaarbrücken, GermanyinternshipJun 2026 – Aug 2026

Summary

CISPA Summer Research (CSR) Internship Program

What I did

  • Conducted independent research in Dr. Ana-Maria Cret¸u's group on reconstruction attacks that recover individual-level records from aggregate statistics.
  • Built reconstruction pipelines in Python using PuLP, CBC and Gurobi, modelling the problem as Integer Linear Programming over aggregate constraints.
  • Implemented custom second-solution and uniqueness checks using integer cuts and Big-M formulations, and investigated solver behaviour, k-way marginals and limited-query reconstruction.
  • Reproduced findings from existing papers on database reconstruction attacks and investigated factors influencing reconstruction success.
  • Developed an ILP-based reconstruction pipeline in Python and experimented with different combinations and dimensions of aggregate statistics.
  • Investigated whether successful reconstruction necessarily meant the solution was unique.

Results

  • Achieved 100% reconstruction accuracy on a synthetic dataset of 10 records using specific combinations of 4-way marginals.
  • Identified how the amount, dimensionality, and overlap of released aggregate data affect reconstruction accuracy.
  • Tested solver behavior across 10,000 random seeds.
PythonPuLPCBCGurobiInteger Linear ProgrammingBig-M formulations

Teaching Assistant

Lahore University of Management SciencesLahore, Pakistanpart timeAug 2024 – Dec 2025

Summary

Key Responsibilities: • Teaching Assistant for CS100 Computational Problem Solving, CS200 Object-Oriented Programming, and CS331 Artificial Intelligence & Machine Learning; supported 320+ students through assignments, tutorials, code reviews and office hours.

What I did

  • Used automated workflows to build code and run test suites when changes were pushed to GitHub.
Computational Problem SolvingObject-Oriented ProgrammingArtificial IntelligenceMachine LearningGitHub Actions

Software Engineering Intern

Intech Process AutomationLahore, PakistaninternshipJun 2025 – Aug 2025

Summary

AI-Powered Tenant Health & Support Assistant

What I did

  • Built an end-to-end internal tenant-management application using Python, FastAPI, Jinja2 and Azure SQL, centralizing customer license, usage and lifecycle information for support teams.
  • Designed REST-backed workflows, sorting/filtering, pagination, alerts and a custom weighted tenant health scoring engine to identify expiring, inactive and high-potential tenants.
  • Integrated Azure OpenAI (GPT-3.5 Turbo) to generate context-aware trial-ending, upgrade and re-engagement emails directly from tenant metadata.
  • Migrated persistence from CSV files to Azure SQL using SQLAlchemy/pyodbc and deployed the application to Azure App Service, completing requirements, development, testing, documentation and handover.
  • Consolidated tenant information to automatically calculate health scores and identify accounts requiring attention.
  • Integrated Azure OpenAI to generate personalized customer outreach emails.
  • Deployed a full-stack application on Azure App Service and managed an Azure SQL Database.
  • Configured cloud services and environment variables.

Results

  • Built and deployed an end-to-end system that automated tenant monitoring, prioritization, and AI-assisted communication.
  • Designed system to support a portfolio of hundreds of tenants.
  • Estimated potential 60–80% reduction in drafting time for customer outreach emails.
PythonFastAPIJinja2Azure SQLRESTAzure OpenAIGPT-3.5 TurboSQLAlchemypyodbcAzure App Service

Skills

Technical

PythonPython
Android Studio
Agile
FastAPIFastAPI
Integer Linear Programming
FirebaseFirebase
Java
Jinja2
NLP
SQLAlchemySQLAlchemy
PyTorchPyTorch
Azure SQLAzure SQL
Azure OpenAI
FigmaFigma
BeautifulSoup
Firebase FirestoreFirebase Firestore
Atomic database operations
Cloud computing fundamentals
GitHubGitHub
Operating Systems
Computer Systems
GitHub ActionsGitHub Actions

Projects

Campus Event Discovery & Management Platform

Developeracademic
Developed a multi-role Android application for students, organizers and administrators, supporting campus event discovery, organization management, event approval and attendee registration. Built functionality for event search, RSVP/cancellation, scheduling-conflict detection, QR-based e-tickets, attendance scanning, notifications and role-based administrative workflows. Designed the application around object-oriented model-controller-repository layers using Java, Android Studio, Firebase Authentication, Firestore, Firebase Storage and Firebase Cloud Messaging. Worked in a five-member Agile team using user stories, Kanban, sprint planning, UML, CRC cards, testing, JavaDocs and iterative integration.
JavaAndroid StudioFirebase AuthenticationFirestoreFirebase StorageFirebase Cloud MessagingUMLCRC cards

Urdu News Classification

Developeracademic
Built a complete NLP pipeline and custom dataset of 2,185 Urdu news articles scraped from five Pakistani news platforms using Python, Requests and BeautifulSoup. Implemented Urdu-specific preprocessing including normalization, tokenization, lemmatization and stop-word removal, followed by Bag-of-Words vectorization. Implemented Multinomial Naive Bayes and regularized Logistic Regression from scratch, and trained a PyTorch feedforward neural network that achieved approximately 98.9% accuracy.
PythonRequestsBeautifulSoupPyTorchMultinomial Naive BayesLogistic Regression

Social Link - Student Collaboration Platform

Designeracademic
Designed a unified communication and task-management platform for university societies to reduce fragmentation across WhatsApp, Slack, Notion and other disconnected tools. Conducted user interviews and focus-group research, designed the low-fidelity prototype and interaction workflows, and contributed to iterative usability evaluation and refinement of the final Figma prototype.
Figma

Education

B.S. Computer Science

Lahore University of Management Sciences (LUMS)Computer ScienceGPA 3.73Aug 2022 – Jun 2026

Coursework

Computational Problem SolvingObject-Oriented ProgrammingArtificial Intelligence & Machine LearningSoftware EngineeringHuman-Computer Interaction

Societies & activities

Dean's Honour List

Recognition

Awards

Graduated with Distinction

Lahore University of Management Sciences (LUMS)2026

Dean's Honour List

Lahore University of Management Sciences (LUMS)

Awarded throughout undergraduate studies

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

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

More people on Kanvis

Browse more peopleCreate your own Kanvis