DEVANSHU AGRAWAL

ACADEMIC PROFILE
Degree/Certificate Institution Percentage/CGPA Year
B-Tech Electronic Engineering IIT (BHU), Varanasi 9.38 2024
CBSE (XII) Kanha Makhan Public School, Mathura 94.80 2019
CSBE (X) Kanha Makhan Public School, Mathura 95.00 2017

SKILLS
Programming Languages: C/C++, Python
Areas of Interest: Data Structures and Algorithms, Competitive Programming, Deep Learning, Computer Vision
INTERNSHIP/TRAINING
Software Engineering Intern | Jaguar Land Rover May 2023 - July 2023
Resolved inconsistencies in cruise control simulation, developed and integrated speed limiter feature in simulation.
Incorporated, developed logging wrapper into MATLAB/Simulink by linking header files and shared library files.
Formulated comprehensive test cases using GoogleTest for software production-level unit testing of cruise control
features, also designed user-friendly buttons for cruise control feature GUI.
Exposure: MATLAB, Simulink, C++, GoogleTest
PROJECTS
Handwritten Emoji Captcha Detection | Mosaic'22 Mar 2022 - Apr 2022
Used image processing techniques to segment different characters present in a captcha and recognized each using a
custom CNN based model trained on the custom dataset of 26 classes with accuracy of ~97.63% on training data.
Added features to handle rotated characters, also applied image processing techniques like Adaptive thresholding,
Morphological operations etc. to remove noise present in image. Exposure: OpenCv, Python, Keras, CNN | Github
Pool Detection | Mosaic'22 Mar 2022 - Apr 2022
Implemented a robust model to detect the swimming pools in unconstrained satellite images, addressing occlusion
and other spatial variations. Used bounding boxes for precise pool localization in the images.
Trained YOLOV3 model built upon DarkNet53 architecture on dataset images with PASCAL VOC annotations, got
MSE metric average loss ~0.2215 on training data. Exposure: OpenCv, Python, Keras, CNN, YOLOV3 | Github
Stained Cell Count | MLWare'22 Feb 2022 - Mar 2022
Implemented a regression model to count the stained blood cells present in the pigmented test sample image.
Trained a separable CNN architecture model named Xception on the images dataset with cell count as label, got
MSE average metric loss ~0.3467 on training set. Exposure: Python, Keras, CNN, DNN | Github
POSITION OF RESPONSIBILITY
Training And Placement Representative | Training And Placement Cell, IITBHU Apr 2022 - Present
Integral part of 50+ member team coordinating internships and placements for thousands of IITBHU students.
Liaised with global HR professionals, and successfully led inaugural editon of Career Development Program 2023.
Under-Secretary-General for Social Media | IITBHU MUN 2022 Jan 2022 - Mar 2022
Significantly elevated social media reach by over 110% through pivotal leadership and strategic initiatives.
Engaged global organizations, securing 250+ delegates from 15+ nations across 4 continents for MUN conference.
HONOURS AND ACHIEVEMENTS
1st Runner-up among 40+ teams in Mosaic'22, an event based on Computer Vision at Udyam'22.
Expert on Codeforces max rating 1682 (ID: devanshu-08), 5 Star on CodeChef max rating 2026 (ID: dnag88).
Guardian on LeetCode with max rating 2183 (ID: devanshu-08).
Secured Global Rank 485 in Google Kick Start Round D 2022 and Global Rank 238 in Codeforces Round 891.
Secured Global Rank 24 in International Coding Marathon'22 held on CodeChef during Technex'22.
Secured Global Rank 79 in Newton's Grand Coding Challenge 2022.
Ranked among the Top 5 teams in Machine Learning based event MLWare’22 during Technex’22.
Ranked among the Top 10 teams in Data Science-based hackathon held at Cassandra'22 during Udyam’22.
Achieved AIR 1677 in JEE Advanced 2020 and AIR 3204 in JEE Mains 2020.

T: 7906232338 E: [email protected] Address: 1412, Gali Megha, Chowk Bazar, Mathura