laiba.khan
Available for internships & entry-level roles

AIsystemsthatlearn,see,andreason.

I'm Laiba Khan, an artificial intelligence undergraduate at NED University, building in

Machine Learning
View projectsGet in touch
Scroll

Profile

About

01

Artificial Intelligence undergraduate with strong foundations in machine learning, computer vision, data processing, and software development. Experienced in building AI-driven applications using Python, C++, and SQL, with hands-on work in machine learning models, database systems, and low-level programming. Seeking internship or entry-level opportunities in Artificial Intelligence, Machine Learning, Data Science, or AI-driven software development.

Based in
Karachi, Pakistan
Focus
AI / ML / CV
Status
Open to work

Education

07/09/2023 — Current

BS Artificial Intelligence

NED University of Engineering and Technology

Karachi, Pakistan

Final grade 3.504 / 4.00

Matriculation & Intermediate

Pre-Medical

Mari Energies Higher Secondary School

Daharki, Pakistan

Toolkit

Skills

02

Programming

C/C++PythonSQLPrologx86 Assembly

Tools

Visual StudioPyCharmJupyter NotebookExcelAccess

Soft Skills

TeamworkManagementCommunicationComputational ThinkingAdaptabilityProblem Solving
C/C++ /Python /SQL /Prolog /x86 Assembly /Visual Studio /PyCharm /Jupyter Notebook /Excel /Access /Teamwork /Management /Communication /Computational Thinking /Adaptability /Problem Solving /C/C++ /Python /SQL /Prolog /x86 Assembly /Visual Studio /PyCharm /Jupyter Notebook /Excel /Access /Teamwork /Management /Communication /Computational Thinking /Adaptability /Problem Solving /

Selected Work

Projects

03
01Featured

Roman Urdu Symptom Triage Classifier

Clinical NLP · Fine-Tuned Transformers

A clinical NLP system that predicts triage urgency and body-system category from free-text Roman Urdu symptom descriptions — the script Pakistani patients actually type into a phone. Built solo across six disclosed rounds of error analysis and architecture fixes, with every claim backed by the script that produced it.

  • Fine-tuned XLM-RoBERTa and MuRIL transformers behind a hybrid dual-model serving architecture
  • Iterative error analysis lifted emergency-case recall from 0.27 to 0.97 and body-system accuracy from 0.06 to 0.88 across 6 rounds
  • Hand-verified 1,600+ labels via Cohen's kappa audits and a custom Next.js + FastAPI review tool
  • Deployed with a live demo on Vercel and model weights hosted on Hugging Face Hub
PythonPyTorchTransformersXLM-RoBERTaNLPFastAPI
CodeLive demo
02

Cyber Incident Detection System

Machine Learning & Graph Analysis

A machine learning system that detects cybersecurity incidents from event log data, correlating related security events to surface insider threats.

  • Applied Random Forest classification to flag suspicious user activity
  • Graph-based analysis to correlate related security events
  • Feature engineering & preprocessing to lift model performance
PythonRandom ForestGraph AnalysisSecurity
03

Manipulation Pattern Detection

Deep Learning

A Deep Neural Network that detects and classifies manipulation patterns in conversational data, with a focus on model robustness.

  • Built and tuned a DNN classifier for conversational text
  • SHERLOCK-based verification for stability & output-range checks
  • Evaluated via metrics, visualizations, and robustness testing
Deep LearningNLPModel Evaluation
04

AI-Powered Blog Mining & Comment Generation

NLP & Generative AI

An automated pipeline that mines blog content, scores sentiment and tone, and generates AI-written comments.

  • Web scraping + NLP pipeline for sentiment, tone, and motive inference
  • Visual analytics dashboards with Matplotlib and Word Cloud
  • Hugging Face Transformers + Microsoft Phi-3 Mini for comment generation
NLPHugging FacePhi-3 MiniWeb Scraping
05

Sentiment Analysis Tool

C++ (OOP-Based)

A sentiment analysis application built from the ground up in C++ using object-oriented design.

  • Text-processing algorithms classify positive / negative / neutral
  • Modular class design for scalable, maintainable software
C++OOPNLP
06

Employee Management System

SQL — DBMS Project

A relational database system for managing employee and department records with a focus on data integrity.

  • Schema design with full database normalization
  • Views, triggers, and advanced queries for reporting & auditing
SQLDatabase DesignDBMS
07

Hospital Bed Allocation System

Greedy & Dynamic Programming

A patient-scheduling simulation that allocates limited hospital beds under urgency-based prioritization.

  • Compared Greedy vs. Dynamic Programming allocation strategies
  • Benchmarked execution time, scalability, and allocation efficiency
  • Simulated emergency scenarios for real-time decision-making
AlgorithmsDynamic ProgrammingSimulation
08

Hangman Game

x86 Assembly — Irvine32

A text-based Hangman game built directly in x86 Assembly to strengthen low-level programming fundamentals.

  • Random word selection, input validation, and scoring
  • Hands-on work with registers and memory addressing
AssemblyIrvine32Low-Level

Continuous Learning

Certifications

04
Coursera· 2026

Computer Vision and Speech Recognition

  • Image processing with OpenCV
  • Edge detection, filtering, thresholding, object detection
  • Speech & audio processing using MFCC and spectrograms
  • End-to-end pipelines for AI applications
Google & Coursera

Google AI Professional Certificate

  • 7-course certification: AI fundamentals, prompt engineering, AI-assisted research, content creation, data analysis, and AI application development
  • Generative AI tools, AI-powered workflows, and responsible AI usage

Beyond the Classroom

Leadership & Languages

05
Member — AI Club, NED

Participated in workshops, technical discussions, and AI-focused collaborative learning sessions.

Participant — Inter-University Coding Competitions

Solved algorithmic and logic-building problems using Python, SQL, and C++.

Public Speaker — Seminar on AI

Delivered a presentation on AI fundamentals and real-world applications to students and faculty.

Languages

UrduNative
EnglishFluent
TurkishConversational

06 / Contact

Let's build something intelligent together.

Open to internships and entry-level roles in AI, ML, and data science. Reach out — I reply quickly.

Email

laibasalman789@gmail.com

LinkedIn

laiba-khan

Phone

+92 335 6343466