Scribd
Upload
44 views2 pages

Karun Sharma Resume

Uploaded by

FlicK
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
44 views2 pages

Karun Sharma Resume

Uploaded by

FlicK
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 2

Software Developer

+917973704258 karunsharma1920@gmail.com https://github.com/karu03 Phase-XI, SAS


Mohali, Punjab, 140308

Summary
Results-driven Software Developer specializing in AI and machine learning solutions. Proficient in Python, TensorFlow,
and Keras with experience in full-stack development using Django and Flask. Proven track record of delivering
scalable applications on AWS, focusing on innovative solutions that drive business impact.

TECHNICAL SKILLS
 Programming Languages: Python (Advanced), C++ (Intermediate), JavaScript (Intermediate)
 ML/AI: PyTorch, TensorFlow, Scikit-learn, NLTK, Transformers
 Web Development: Django, Flask, RESTful APIs
 Cloud & DevOps: AWS (EC2, S3, Lambda), Docker
 Data Analysis: Pandas, NumPy, Jupyter Notebooks
 Tools: Git, VS Code, Gradio, Streamlit
Projects
Hindi2Kangri NMT Model Using NLLB 200 Distilled 600M
03/2024 -10/2024

• Engineered a fine-tuned Meta NLLB-200 model reducing test loss by 40% (0.2621 to 0.1573)
• Fine-tuned Meta’s NLLB 200 model for a low-resource language using PyTorch and AMP (Automatic
Mixed Precision).
• Evaluated translation quality with BLEU and ROUGE metrics, achieving strong performance in low-
resource NLP tasks.
• Technical Challenge Solved: Overcame limited training data by implementing innovative data
augmentation techniques, increasing dataset size by 3x
Person Re-Identification Project
05/2023- 10/2023

• Developed an advanced deep learning model achieving 93.24% Rank-1 accuracy and 80.78% map on
Market-1501 dataset
• Implemented Part-based Convolutional Baseline (PCB) architecture, improving model robustness for
various viewing angles
• Optimized training with linear warm-up, float16 precision, and DDP, reducing training time by 65%
• Impact: System deployed in a pilot program at local shopping mall, improving security response time
by 40%

ML-NN-AND-AI-MODEL-q-a
11/2022 - 04/2023

• Built a comprehensive Q&A bot, Integrated with OS performs basic I/O tasks
• Implemented voice command functionality using Web Speech API, increasing accessibility
• Technical Challenge: Developed custom caching mechanism reducing response time by 60%

Education
Govt. Degree College, Sarkaghat
Bachelor of Computer Applications
GPA: 7.92/10 06/2020 - 05/2023
• Coursework in Algorithms, Computer Programming, Core concepts like: OS, System Design and DBMS
• Member of Badminton Club
Languages
English Hindi
Certification
Advanced Learning Algorithms - Stanford University
08/2022-09/2022
Supervised Machine Learning: Regression and Classification - Stanford University
09/2022 –10/2022
Python for Data Science, AI & Development - IBM
08/2022-10/2022

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy