Fizzah Amir

Backend Engineer & ML Enthusiast | Building Scalable Systems from Redis Clones to RAG Pipelines

Based in: Lahore, PakistanMost recently: Core Team Member, Webmaster, IEEE ITU Student Branch

Computer Science student at ITU with a focus on high-performance backend systems and machine learning. Proven track record of building complex systems from scratch, including a Redis clone and RAG-based document assistants. Experienced in processing large-scale data (2M+ records) and deploying robust APIs on AWS.

Experience

Core Team Member, Webmaster

IEEE ITU Student BranchLahore, Pakistanwork

Summary

Serve as Webmaster on the IEEE core team, managing and maintaining the society's web presence. Built a chatbot for IEEE Orientation that answers student queries in real time, helping new students get instant information during orientation.

Results

Developed an orientation chatbot to replace traditional society introductions, providing real-time answers to student queries.
Successfully handled queries for 50-55 students during the IEEE orientation, providing instant information and reducing manual workload for the core team.

PythonChatbot DevelopmentWeb Development

Skills

Technical

NumPy

Python

Pandas

C++

DSA

OOP

Random Forest

Feature Engineering

Hugging Face Sentence Transformers

RAG Pipelines

LLM Prompting

REST APIs

TensorFlow

Scikit-learn

XGBoost

Logistic Regression

Decision Tree

PostgreSQL

SQL

Django

Django REST Framework

JWT Auth

Git

Dart

FAISS

SMOTE

Keras

Linux

AWS EC2

OpenCV

Time-Series Analysis

CNN

RBAC

Flutter

Oracle Database

GRU

Optical Flow

Railway Cloud

scikit-image

Express.js

Node.js

Projects

Insightly-PDF - RAG-Based Document Q&A Assistant

Developerpersonal

Built a full-stack Retrieval-Augmented Generation assistant that lets users upload PDFs and ask natural-language questions, returning grounded, citation-backed answers from the document content. Developed a Django REST backend to handle PDF ingestion, text extraction/chunking, embedding generation, and query orchestration through a RAG pipeline. Indexed document embeddings in a vector store for semantic retrieval, feeding the most relevant chunks into an LLM prompt to generate accurate, context-grounded responses. Built a React frontend for PDF upload, chat-style Q&A interaction, and displaying source-linked answers, communicating with the backend via REST APIs.

PythonDjangoDjango REST FrameworkReactRAGLLM PromptingVector SearchREST APIs

Digital Wallet System API

Developerpersonal

Built a complete digital wallet REST API (similar to EasyPaisa/JazzCash) with Django REST Framework and PostgreSQL, deployed on AWS EC2. Implemented JWT authentication with custom claims, Argon2 password hashing, and Role-Based Access Control across 3 roles. Designed 29 REST API endpoints using raw SQL with explicit transactions and parameterized queries; wrote 8 PostgreSQL triggers and a 5-level automatic rank system.

PythonDjangoDjango REST FrameworkPostgreSQLJWTAWS EC2Argon2SQL

Visit the site

AeroHub Management System

Developeracademic

Full airport management system in C++ with a custom HTTP server, deployed on Railway Cloud with an HTML/CSS/JS frontend. Implemented 6 data structures from scratch: B-Tree, Dijkstra's algorithm on a graph, HashMap, Linked List, Priority Queue, and Bitmap.

C++Custom HTTP ServerHTMLCSSJSREST APIRailway CloudFile I/O

Visit the site

In-Memory Cache Server - Redis Clone

Developeracademic

Built a single-threaded in-memory key-value server in C++ implementing the Redis wire protocol (RESP) with an event-driven epoll core. Built four core data types from scratch and added crash-safe persistence; sustained 50,000+ SET/GET ops/sec in benchmarks.

C++epollTCP SocketsRESP ProtocolSkip ListsHash TablesSystems Programming

Flight Management System

Developeracademic

Console-based C++ system with flight scheduling, seat selection, reservation management, and pilot assignment using core OOP principles and the Singleton pattern.

C++OOPFile I/OSingleton Pattern

Color Detection App for Color Blind Users

Developerpersonal

Flutter mobile app enabling color blind users to identify colors in real time with an accessible interface.

FlutterDartMobile Development

Meeting Transcriber & Q&A Assistant (RAG)

Developerpersonal

Built a meeting transcription and question-answering tool converting recorded audio into searchable, queryable knowledge via a Retrieval-Augmented Generation (RAG) pipeline. Transcribed audio to text, chunked transcripts, and generated dense vector embeddings using a Hugging Face Sentence Transformer (all-MiniLM-L6-v2) for semantic search. Indexed embeddings in FAISS to retrieve the most relevant transcript chunks per query using cosine similarity. Implemented a RAG pipeline feeding retrieved context into an LLM prompt for grounded, citation-backed natural-language answers, plus automatic summarization and action-item extraction.

PythonHugging Face Sentence TransformersFAISSRAGLLM PromptingSpeech-to-Text

Electricity Theft Detection Using Machine Learning

Group Memberpersonal

Built an end-to-end ML pipeline to detect electricity theft from raw smart meter time-series data (UCI dataset, 2M+ records). Performed full data preprocessing: missing-value handling via forward-fill and interpolation, noise removal, normalization, and daily/weekly aggregation. Engineered features including rolling average consumption, peak vs. off-peak ratios, consumption trends, variance, and hour/day-wise behavior patterns. Trained and benchmarked classifiers - Logistic Regression, Decision Tree, Random Forest, XGBoost, and a TensorFlow-based neural network - addressing class imbalance using SMOTE. Produced alerts, classification reports, and consumption-pattern visualizations as post-processing output.

PythonTensorFlowScikit-learnXGBoostPandasNumPySMOTEMatplotlib

Education

BS Computer Science

Information Technology University (ITU)Computer Science2024 – 2028

Coursework

OOPDSAAlgorithmsNetwork ProgrammingAdvanced Database Management

Societies & activities

IEEE ITU Student Branch

FSc, Pre-Engineering

Kinnaird College for Women (KCW)Pre-EngineeringOct 2022 – Aug 2024

High School Diploma

Beaconhouse School SystemMathematics, Physics, CSJul 2022 – 2024

Recognition

Awards

Competitive Programming - SOFTEC

SOFTEC2026

Solved algorithmic and data structure problems in a competitive environment.

Competitive Programming - CodeRush

CodeRush2025

Competed in algorithmic problem-solving challenges, applying data structures and algorithms under time constraints.

Certifications

Foundations of Data Science

GoogleCredential VAQAXU7QH6MU2026

Cybersecurity Bootcamp: ISO 27001:2022

Coursera2025

Meta Backend Developer Professional Certificate

The Power of Statistics

Google

Go Beyond the Numbers: Translate Data into Insights

Google

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

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

More people on Kanvis

Browse more people Create your own Kanvis