Labib Asari

Download as pdf or txt
Download as pdf or txt
You are on page 1of 1

LABIB ASARI

Ó+(91) 9967547973 ∣ [email protected] ∣ ¯LinkedIN ∣ ‡Github


EDUCATION
Veermata Jijabai Technological Institute, Mumbai Dec 2021 - Present
Second Year B.Tech in Computer Engineering. CGPA - 8.9
St Xaviers 2019 - 2021
HSC - 97.1% ∣ MHT-CET - 99.9%ile ∣ JEE-Mains - 98.1%ile
EXPERIENCE
Contributor, GNUAstro - Google Summer Of Code
CPython, Numpy C-API, CUDA, Build Systems Feb 2023 - Present
• Integrated CUDA and OpenCL with Gnuastro build system for GPU support, optimized convolution by 400x.
• Developing a Python Package for Gnuastro using the C-Python API in Python.
• Implemented an error recording and reporting mechanism for Python and C.
• Designed new Data objects in Python for representing Gnuastro’s core data structures using Numpy C-API.
Finalist, Eyantra Robotics Competition - IIT Bombay
Functional Programming, Elixir, Image Processing, Algorithm Development Oct 2022 - Mar 2023
• Designed a Pothole detecting and filling roadbot using Raspberry Pi and Functional Programming. Imple-
mented parallelization to utilize all cores.
• Implemented Shortest Path, Optimal resource allocation Algorithms. Performed Morphological Operations to
detect target objects.
PROJECTS
DocumentGPT
LLMs, LangChain, FastAPI, Image Processing Mar 2023 - Apr 2023
• Created a web-app to connect LLMs with external knowledge bases to deliver accurate, up-to-date information
having the correct context at hand.
• Developed an API for accepting multiple sources of data such as Research papers, docs blogs, books, websites
and videos.
• Added support to chose between OpenSource(HuggingFace) and Paid(OpenAI) LLM models in the backend.
No-Code DL
Deep Learning, PyTorch, NextJS, ReactFlow, Flask Nov 2022 - Feb 2023
• Enabled users with no coding experience to create unique models using a drag and drop functionality or to
use a standard model to address their problems.
• Added support for Standard NNs, CNNs, RNNs and to customize Datasets, preprocessing, hyperparameters.
• Developed a REST API with Flask which accepts model data and returns PyTorch code for the same.
Neural Style Transfer
CNNs, TensorFlow, Data Preprocessing Aug 2022 - Oct 2022
• Extracted feature maps of content and style images using Transfer Learning with a CNN (VGG-19).
• Optimized by chosing different layers, tuning several hyperparameters.Added a bash script for access.
• Trained a CycleGAN to perform unpaired image2image translation. to interconvert images of 2 classes.
ACHIEVEMENTS
• Specialist at Codeforces. ● 4 Star at Codechef. ● Knight at Leetcode (Top 2%).
• Listed as an Open-Source Contributor at OpenCV
• Secured 4th rank Globally at Eyantra Robotics Competition IITB 2023 among 376 teams.
• Hackathons : Winners Smart India Hackathon,VJTI intranary. Finalist Code Odyssey 2.0. and Clash of Codes.
• Coding Events : Winner CheckpointHacker, CodingSurfers @Technovanza, Algocon @SRA, CodeEscape @CoC.
POSITIONS OF RESPONSIBILITY
• CoC VJTI : Organized CodeHunt - a Competitive Programming, workshop did problem setting and testing.
• Leading The CP club, promoting problem solving and algorithmic thinking among peers and juniors.
• Organized introductory workshop in C and embedded systems for juniors at SRA VJTI.

You might also like