Antreev Singh Brar

4th Year Undergraduate — Department of Computer Science and Engineering

R [email protected] | Ó +91-7589244907 | ‡ antreev-brar | ¯ antreevbrar

Academic Qualifications
Year Degree/Certificate Institute CPI/%
2019 - Present B.Tech in Computer Science and Engineering Indian Institute of Technology, Kanpur 9.7/10
2019 Class XII (CBSE) LRS DAV Sr. Sec. Model School 92%
2017 Class X (ICSE) Assumption Convent School,Abohar 94.8%

Achievements & Awards

• Awarded Aditya Birla Scholarship, Amongst top 15 CSE candidates across all IITs
• Awarded Academic Excellence Award, IIT Kanpur for perfect 10 CPI in 1st year
• Awarded Academic Excellence Award, IIT Kanpur for exceptional performance in the academic year 2020-21
• Awarded Prof.JN Kapur Prize, IIT Kanpur for the best Second year undergrad in the Math courses
• Achieved All India Rank 119 in IIT JEE-Advanced 2019 out of 245,000 candidates
• Achieved All India Rank 179 in IIT JEE-Mains 2019 out of 1.3 million candidates
• Achieved All India Rank 72 in KVPY 2019, Conducted by IISC and Gov. of India
• Among top 1% in NSEC & NSEA 2019, and qualified for InChO & INAO conducted by HBSCE in 2019
• Awarded NTSE Scholarship by the Gov. of India for being amongst top 1000 students across the nation
• Papertowns: An AR product approved under the Pre-seed Ideation Program, IIT Kanpur
Work Experience
Tower Research Capital May’22 - Jul’22
Limestone Team, Quant Researcher Intern
• Built statistical models to generate long-range alphas for an aggressive trading strategy using market signals
• Employed various sampling and feature selection techniques to efficiently use extensive historical data and collect stock agnostic
features for enhanced model generalization
• Experimented with various modelling techniques- NGBoost, XGBoost, Decision tree-based models and developed architectures
to capture prediction uncertainty and confidence intervals
• Designed effective de-noising pipelines to realize improved predictions and PnL scores
Prior-free Strategic Multiagent Scheduling Jul’20 - Nov’20
Research Project under Prof.Swaprava Nath n Research Paper
• Implemented an effective priority-based scheduling algorithm to maintain social distancing among visitors in a crowded store
• Proposed mechanism transformed the user-preference matrix into Linear Programming problem and assigned delay (payment) via
Vickery-Clarkes-Groves (VCG) rule
• Utilized the novel approach in the backend development of Mobile application Doori
Sogage May’21 - Jul’21
Software Engineering Intern
• Implemented a data-fetcher using Django REST framework to publish user’s account data in Google Cloud Pub-Sub
• Designed and implemented a robust data pipeline using Apache Beam to transform streaming data and store it in Google Big
Table. Also designed a common schema to harmonise data from all social media networks
Projects
Building GemOS - Operating Systems Aug’21 - Nov’21
Course Project, Prof. Debadatta Mishra
• Created file archiving utility and enabled IPC using system calls like pipe(), fork() and exec()
• Implemented system calls for pipe and persistent pipe structures sharing order preserved data concurrently
• Developed a series of threading APIs with private memory areas by extending clone() system call
• Built a debugger for functions using INT3 featuring stack backtrace of functions
Golang to MIPS Compiler - Compiler Design Jan’22 - Apr’22
Course Project, Prof. Amey Karkare & Prof. Subhajit Roy ‡Project
• Implemented a Golang compiler from scratch with assembly language generated for MIPS64 bit architecture
• Designed a lexer, parser and semantic analyzer that supports Go features including Multidimensional Arrays, Short Variable
Declaration, Multilevel Pointers, Structs, Floats , Strings and Switch statements.
• Implemented Advanced features such as automatic type inference, short-circuit evaluation, compile-time constant expressions
evaluation, pointers, multi-level break, recursion, string manipulation functions and basic features like conditionals, loops, functions
and warning generation for erroneous programs.
• Awarded with the best score for the project among all groups in the course.
Adversarial Training is All You Need Jan’22 - Apr’22
Course Project, Prof. Priyanka Bagade ‡Project
• Examined the defense of Adversarial Training against Poisoning attacks by implementing the defense against BadNet-type
backdoor attack and Clean Label backdoor attack in Keras
• For clean label attack, the method increased the test accuracy on images with backdoor from 1.39 % to 90.68% on MNIST
• For Badnet attacks,the method increased the test accuracy on images with backdoor from 1.17% to 91.71% on MNIST and from
8.64% to 54.86% on CIFAR-10
GenCoin/IITkcoin May’21 - Jul’21
Programming Club IIT Kanpur ‡ Gen-Coin
• Developed a Centralised pseudo-coin system in GoLang for IITK Campus with net/http library for client/server implementation
• Involved Database Management, User Authentication with Hashing & Salting and User Validation with JSON Web Tokens(JWTs)
using bcrypt and jwt-go packages
• Explored and implemented various techniques such as concurrency and deadlock prevention to keep transactions safe in SQLite
engine and maintain ACID properties in the SQL database
IITKBucks Apr’20 - Jul’20
Programming Club IIT Kanpur ‡ IITKBucks
• Developed my own client for a custom Blockchain based Cryptocurrency using NodeJS and ExpressJS
• Implemented Multi-Threading for Block-Miner, to run in parallel with server and ngrok to propagate blocks and data among peers
Model Zoo Apr’20 - Jul’20
Programming Club IIT Kanpur ‡Model-Zoo
– Studied Deep Learning research papers and implemented them from scratch using Keras
– Explored domains of Computer Vision , GANs and MultiModal Neural Networks
– Worked with YOLOv3, SS-GAN, ResNet-34 models and CIFAR-10, MNIST, and Flickr8k datasets with max acc. of 91.8%
Speech Ledger Oct’20 - Nov’20
FlipkartGrid 2.0 ‡ Speech-Ledger
• Developed a custom pipeline for Background noise suppression, Speaker Segmentation & Diarization, and voice-to-text translation
• Fine tuned Dual-Signal Transformation LSTM Network for Real-Time Noise Suppression, Text-independent Speaker recogni-
tion module based on VGG-Speaker-recognition as Encoder and Speaker diarization based on UIS-RNN as Decoder to work
synchronously with Flipkart’s voice-to-text API
Competitive Programming
• Codeforces : [Handle: wolfiee] Maximum rating of 1929 (Candidate Master)
• Google Kickstart 2021 : Global Rank 231 in Round D and 488 in Round C among more than 10,000 participants
• ICPC 2021 : Achieved Rank 91 in Preliminary and 308 in Amritapuri regional round out of 7000+ teams
Technical Skills
Proficient: C • C++ • Golang • Python • Javascript
Web Development: NodeJS • Django • Express • HTML
Utilities:Git • Docker • Kubernetes • SQLite • Apache Beam • TensorFlow • Matplotlib •LATEX• Keras • Bash • Vim
Exposure: ReactNative • Verilog
Relevant Coursework
Data Structures and Algorithms Compiler Design Operating Systems
Advanced Algorithms Computer Network Computational Genomics
Linear Algebra and Ordinary Diff. equations Multivariate Calculus Deep Learning for Computer Vision
Mathematics for Computer Science Logic in Computer Science Image Processing∗
Derivative Contracts∗ Intro to ML∗
* - Ongoing Courses

Positions of Responsibility
• Secretary, Programming Club IIT Kanpur
Organised various Pclub events throughtout the year
• Teaching Assistant, Data Structures and ALgorithms
Tutored Data structure course to 2nd year Undergrads at IITK.
• Secretary, Academics and Career Council IIT Kanpur
Coordinated a blog series, and organised various events of AnC
• Student Guide, Counselling Service IIT Kanpur
Guiding and Leading 6 freshmen into college life at IIT Kanpur