Saad Iftikhar
Islamabad, Pakistan
Summary
Aspiring AI practitioner with a strong passion for using artificial intelligence to solve real‑world problems and enhance efficiency. Fascinated by
the potential of AI to simplify complex tasks and accelerate innovation, I am eager to apply my knowledge in machine learning and data‑driven
problem‑solving to impactful projects. Committed to continuous learning and creative thinking, I seek an internship opportunity where I can
contribute to building intelligent systems that make work smarter and faster.
Education
FAST NUCES(National University of Computer and Emerging Sciences)
B.S. iN DATA SCiENCE
Military College Jhelum
F.SC (PRE‑ENGiNEERiNG)
Islamabad, Pakistan
Aug. 2022 ‑ Current
Sarai Alamgir, Jhelum
May. 2016 ‑ Jul. 2021
Experience
FiVER
Feb, 2022 ‑ May, 2022
• Worked as Virtual Assistant to manage client’s Amazon Account. From making the listing to running PPC Ad campaigns.
Projects
STREET FiGHTER AI‑POWERED BOT
Dec, 2024
• Developed an AI‑driven bot using PyTorch and a Multi‑Layer Perceptron (MLP), integrated with BizHawk emulator via sockets for real‑time game
state analysis and automated control inputs.
• Designed a robust ML pipeline in Python, leveraging pandas and scikit‑learn for data preprocessing, logging gameplay data to CSV, and opti‑
mizing model performance with batch normalization, dropout, and early stopping.
• Built a real‑time ETL pipeline to extract, enrich, and load incomplete transactional data from data sources into the DW using master data for
enrichment.
• Engineered a modular codebase, implementing object‑oriented classes for game state parsing, player controls, and JSON command serializa‑
tion, enabling scalable data collection and dual‑player perspective support. (view code).
MUSiC RECOMMENDATiON SYSTEM
July,2024
• Built a scalable music recommendation system using the FMA dataset, providing personalized song suggestions based on user preferences and
listening history.
• Implemented ETL pipeline in PySpark; applied audio feature extraction (MFCC, spectral centroid, zero‑crossing rate), dimensionality reduction,
and normalization techniques.
• Deployed real‑time recommendations using Flask and Apache Kafka; used MongoDB for storage and Apache Spark for model training with
collaborative filtering and ANN algorithms. (view code).
NEAR REAL TiME DATA WAREHOUSE FOR CUSTOMER BEHAViOUR ANALYSiS
July,2024
• Developed a near‑real‑time data warehouse (DW) to analyze customer shopping behavior and support data‑driven promotional strategies.
• Implemented an extended version of the Mesh Join algorithm in Java using Eclipse IDE to support stream‑relation joins for ETL transformation.
• Built a real‑time ETL pipeline to extract, enrich, and load incomplete transactional data from data sources into the DW using master data for
enrichment.
• Utilized SQL queries to extract valuable insights from the data stored in the DW, supporting decision‑making and business strategies. (view
code).
Skills
• Programming Languages: Python, C++, SQL
• Data AI Tools: Pytorch, Pandas, NumPy, Scikit‑learn, PySpark, Apache Spark, Apache Kafka, MongoDB
• Web Development: FastAPI, Flask, HTML/CSS, Git, Eclipse IDE
• Concepts: Machine Learning,Neural Networks , Data Structures Algorithms, ETL/ELT Pipelines, Real‑time Sys‑
tems
