Abdul Rehman Junaid

Machine Learning and Full-Stack Engineer building end-to-end AI products.

Based in: Lahore, PunjabCurrently: Software Engineer, Strategic Systems InternationalExperience: 1 year

A high-achieving Computer Science undergraduate with a 3.94 GPA and expertise in building full-stack ML systems from data pipelines to React frontends. He has proven experience in developing predictive forecasting modules, behavioral biometric authentication, and MLOps pipelines using tools like TensorFlow, FastAPI, and Docker.

Experience

Software Engineer

Strategic Systems InternationalLahore, PunjabworkJun 2026 – Present

Summary

Building a full-stack workforce optimization platform with an Angular frontend and a .NET backend, letting businesses schedule shifts, manage staff and plan labour, owning backend services, APIs, data models and the user-facing application. Developed a predictive ML module that forecasts future spending from historical scheduling and payroll data, giving managers a data-driven view of upcoming labour costs. Built an optimization engine that recommends the best shift configurations from a company's historical data or reusable templates, helping businesses cut manual scheduling effort and improve coverage and cost. Integrated an AI chatbot using retrieval-augmented generation (RAG) over an application knowledge base, giving users instant in-product guidance. Built data pipelines to clean and prepare historical data for the forecasting and optimization models, and evaluated model outputs against historical baselines to keep recommendations reliable.

What I did

Evaluated forecasting models by comparing historical data predictions against actual outcomes.

Results

Enabled managers to measure scheduling strategy changes by comparing predicted labor costs against actual end-of-month results.
Developed a forecasting module using a mixture of operations research and machine learning models.
Onboarded 2 businesses from the Middle East to the SaaS product.
Achieved an average decrease of about 10-12% in labor costs during testing against historical data.

Angular.NETMLRAGData Pipelines

Security Operations Center Intern

Systems LimitedLahore, PunjabinternshipJun 2025 – Aug 2025

Monitored SIEM dashboard (Cortex XDR) performing initial triage on high-volume security alerts across diverse environments. Analyzed 50+ daily security logs, effectively filtering false positives by correlating data from firewalls and endpoint logs to maintain high detection accuracy. Identified and mitigated threats such as brute-force attempts and anomalous login activities, protecting critical company infrastructure from damage.

SIEMCortex XDRFirewalls

Business Development Intern

CloudpacerLahore, PunjabinternshipJun 2024 – Aug 2024

Reached out to potential clients about business. Took initial technical calls with clients explaining what the company does and how their work will be done. Used Salesforce and HubSpot to manage client data and automated emailing.

SalesforceHubSpot

Skills

Technical

Python

Machine Learning

TensorFlow

Deep Learning

React.js

FastAPI

Fine-tuning

Model Evaluation (Evals)

MLOps

Data Pipelines

RAG

MLflow

DVC

AI Agents

Docker

SQL

Kotlin

.NET (C#)

Angular

TypeScript

LangChain

AI Automations

Linux

AWS

C++

Pen testing

Kubernetes

Rust

Go Lang

AI Token Optimization

Background jobs

Rapid prototyping

Capture The Flag (CTF)

Languages

English

Projects

ML-based Malicious Domain Detection

Developerpersonal

Built an end-to-end MLOps pipeline that classifies malicious domains from TLS certificate features. Versioned datasets with DVC, tracked experiments, metrics and model comparisons in MLflow, and reached around 90% prediction accuracy. Containerized the inference API with Docker, deployed a live endpoint on Render, and built a React web interface for testing live domains and certificates.

PythonMachine LearningMLflowDVCFastAPIDockerReact.jsRender

Visit the site

Blood Donation App

Full-Stack Developerpersonal

Developed an Android app that shows real-time blood donation requests on a map by user location and blood type. Built the backend on Firebase (real-time data sync, authentication) with request filtering, donor reviews and request acceptance, and a personalized experience from user profiles.

KotlinAndroidFirebaseGoogle Maps API

Sync and Solve

Game Developerpersonal

Developed a 2D multiplayer puzzle-platformer inspired by Pico Park using the Unity game engine. Implemented local co-op mechanics, synchronized player interactions, and level-based challenges using C# scripting and Unity's physics engine.

UnityC#Multiplayer Game Development

Airplane Management System

Full-Stack Developerpersonal

Built a full-stack web application in ASP.NET using the MVC architecture for managing flight schedules, bookings, and passenger data. Implemented CRUD operations with Entity Framework, connected SQL Server for relational data storage, and ensured secure, role-based access control. Designed responsive front-end views using Razor Pages.

ASP.NETMVCSQL ServerEntity FrameworkRazor Pages

Early Fire Detection System

Developerpersonal

Built a real-time fire detection system using custom-trained CNNs in TensorFlow. Handled data preprocessing, trained and evaluated image classification and object detection models with standard evaluation metrics, and triggered automated fire alerts from live input.

PythonTensorFlowCNNComputer VisionDeep Learning

Fine-tuning Accessibility Tool

personal

A tool designed to help non-ML engineers evaluate if fine-tuning is necessary by uploading a CSV, checking for data problems, and comparing a baseline against a fine-tuned DistilBERT model.

FastAPIReactTypeScriptMLflowDocker

Continuous Authentication using Behavioral Biometrics

Lead Developeracademic

Final-year project. Built an end-to-end ML system that continuously verifies a smartphone user's identity from behavioral biometrics, classifying sessions as owner or imposter with a TensorFlow deep learning model and automatically locking sensitive apps. Designed the data collection, feature extraction, training and evaluation pipeline, plus an adaptive retraining loop that fine-tunes the model as the owner's behavior drifts. Served inference through a FastAPI backend to a Kotlin Android client. Achieved 88% accuracy while keeping the false rejection rate low to avoid frustrating legitimate users.

PythonTensorFlowDeep LearningKotlinAndroidFastAPIAI tools

Bluetooth Home Automation

personal

Built a hardware-software integration using Arduino and ESP boards to control room lights and fans via a Bluetooth-connected application.

ArduinoESP boardsBluetooth

Education

BSc in Computer Science

FAST NUCES LahoreComputer ScienceGPA 3.94Aug 2022

Thesis. Continuous Authentication using Behavioral Biometrics

Cambridge A' Level

Lahore Grammar School Senior BoysAug 2020 – Jun 2022

Societies & activities

Robotics Society

Recognition

Awards

Silver Medal in 1st Semester

FAST NUCES Lahore

Bronze Medal in 3rd Semester

FAST NUCES Lahore

Silver Medal in 4th Semester

FAST NUCES Lahore

Gold Medal in 5th Semester

FAST NUCES Lahore

Sliver in 6th Semester

FAST NUCES Lahore

Bronze in 7th semester

FAST NUCES Lahore

Gold in 8th Semester

FAST NUCES Lahore

3rd Position in CS Batch of 2022

FAST NUCES Lahore

FAST NUCES Rector list and Deans Honor list

FAST NUCES Lahore

Certifications

Google Cybersecurity Professional Certificate

Coursera2026

AWS Cloud Foundations

AWS Academy2025

Governance & advisory

President

Robotics Society, Lahore Grammar SchoolotherAug 2020 – Jun 2022

Spearheaded innovative projects, ignited student interest, organized events and developed expertise among peers

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

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

More people on Kanvis

Browse more people Create your own Kanvis