Download The Resume

Muhammad Umar Waseem

umar.waseem@gmail.com | Islamabad, Pakistan | https://muhammadumarwaseem.com

Professional Summary

Muhammad Umar is a Computer Science graduate with hands-on experience in developing and deploying Full Stack Web and Mobile App solutions to cloud providers like Azure and AWS, including Machine and Deep Learning microservices. Muhammad also has experience solving various problems while working with different people as a freelance software engineer.

Experience

Software Engineer

Cowlar Design Studio, Islamabad, Pakistan

  • Collaborated with teams of Web, Mobile, ML, and Embedded System Engineers, integrating embedded systems, IoT devices, low energy devices, Bluetooth systems, online web APIs with mobile and web apps, resulting in seamless and enhanced user experiences.
  • Automated workflows using Github Actions and Gitlab DevOps workflows, ensuring product quality and agility, significantly reducing errors and improving development speed.
  • Optimized product performance and costs for mobile platforms, achieving a 50-60% improvement in efficiency and cost-effectiveness.
  • Worked on initiatives to improve existing team workflows, such as suggesting agile no-code tools and keeping team members aware of new libraries and improvements.

Head App Development

Google Student Developers Club, Islamabad, Pakistan

  • Developed mobile applications using Flutter, providing innovative solutions and contributing to various projects within the community.
  • Conducted over 5 workshops to educate and enable students, successfully familiarizing at least 50% of the attendees with essential software engineering skills and technologies.
  • Led app development initiatives, mentoring peers and fostering a collaborative learning environment to enhance the technical capabilities of the club members.

Freelance Software Engineer

Google Student Developers Club, Islamabad, Pakistan

  • Fixed bugs and made major modifications in existing codebases.
  • Deployed on cloud platforms such as AWS and Azure using DevOps techniques.
  • Worked with a diverse set of problems and use cases.
  • Fulfilled various requirements, including management systems, simple to complex apps like apps handling whole operations of a hotel/restaurant, social media competitors, etc.

Skills

Python - TensorFlow / Keras / PyTorch - Numpy / Pandas - MLflow - Flask / Django / Model Deployments - Docker / Kubernetes / GitHub Actions / Jenkins - Jira / Airflow / C++ / Backend Development / Node.js / Golang

Projects

Retrieval Augmented Gemini Chatbot

  • Uses LangChain, a new framework for working with different AI models
  • Uses Google’s Gemini as a large language model
  • Reduces hallucination and gets the latest data ensuring privacy and concise answers

Airflow Data Extraction Pipeline

  • Seamless ETL pipeline
  • Extracts data from source, performs cleaning, and sends to storage
  • Scheduled and manual pipeline triggers

IslamKGHub Final Year Project

  • Knowledge graph based Q&A system for relevant stakeholders to study and ask about Islamic Hadith, a vast database of knowledge maintained over many years
  • Uses natural language processing to serve user queries
  • Graph DB and SparQL queries to extract data from the knowledge graph
  • Backend system APIs fully deployed and seamless for communication of the graph to frontend

GAN Cifar-10 Generative Image Modeling

  • Generative Adversarial Network trained on CIFAR-10 dataset
  • Compute-intensive model training while saving model checkpoints to continue training in supervision or on a different platform for GPU availability
  • CUDA and CUDNN usage for model training over a range of 200 epochs on an NVIDIA RTX 3060ti

Siamese Networks Image Similarity

  • Siamese neural network for image similarity calculation
  • This model contains a pair of neural networks with shared weights for better performance and similarity calculation

Golang Blockchain

  • A fully fledged blockchain implemented in Golang
  • Peer-to-peer network system for nodes
  • Asynchronous Go routines to handle different processes in the blockchain
  • Adjustable block mining difficulty to simulate the real world
  • Structured Merkle tree made from each transaction and block mined

Find more projects on github.com/Umar-Waseem or on /projects

Education

Bachelors of Computer Science

Fast National University of Computer and Emerging Sciences

2020 - 2024

Subjects: Data Structures, OOP, Artificial Intelligence , Generative AI, MLops

Certifications

  • Azure AZ900
  • Aws Certified Cloud Practitioner (in progress)

Volunteer Work

  • Community Volunteer with Pakistan Red Crescent May
  • 2024