UMAR BALAK
===========

Love building intelligent systems that scale.
Exploring backend systems, cloud architecture, and Go for high-performance services.

---------------
Connect With Me

• Portfolio    -> https://www.umarbalak.me
• LinkedIn     -> https://linkedin.com/in/umar-balak/
• X            -> https://x.com/Umar_Balak
• Medium       -> https://medium.com/@UmarBalak
• Email        -> umarbalak35@gmail.com

----------
Focused On

• Classical Machine Learning
• Deep Learning & Generative AI
• Federated Learning & Privacy-Preserving Systems
• Distributed Backend Architecture
• Databases & Cloud Infrastructure

--------------
Skills & Tools

AI & ML           -> Python, TensorFlow, Keras
                     Scikit-learn, NumPy, Pandas, OpenCV

Backend           -> FastAPI, REST APIs, Async Systems
                     System Design

Databases         -> PostgreSQL, MySQL, MS SQL Server, Pinecone

DevOps & Cloud    -> Linux, Git, Azure, Render, Vercel, Google Cloud

-------------
Selected Work

• VanillaNets
  Description  -> Transparent, NumPy-only neural network library for learning & experimentation
  Link         -> https://pypi.org/project/vanillanets/
  Tech         -> NumPy, Python, Neural Networks, From-Scratch Implementation

• Collaborative RAG Learning Platform
  Description  -> SaaS platform enabling peer-reviewed AI knowledge base
  Link         -> https://cognify.cohernt.xyz
  Tech         -> Next.js, FastAPI, LangChain, Pinecone

• Federated Learning Framework
  Description  -> Decentralized model training with privacy & secure aggregation
  Link         -> https://github.com/UmarBalak/adaptfl_client
  Tech         -> TensorFlow, Differential Privacy, FastAPI

• Image Classification Model
  Description  -> Optimized VGG16-based architecture for CIFAR-10
  Link         -> https://github.com/UmarBalak/Cifar10_VGG
  Tech         -> TensorFlow, CNN

------------------
Currently Building

• AI infrastructure for scalable model deployment
• Data pipelines for distributed systems
• Cloud-native backend services in Go

-----------------------
Thanks for stopping by!

If you are building in AI, distributed systems, or backend engineering,
let’s connect and build something meaningful.