Adarsh Roy ⚪

Adarsh Roy

(he/him)

Backend Engineer @ ML and Apps Platform

Dragonfruit AI

About me

Backend engineer with a background in mathematics and computing. I work across ML and apps platforms, from distributed systems and cloud infrastructure to deep learning and multimodal AI projects. I enjoy writing clean, reliable code and turning research ideas into practical applications.

Resume

Education

B.Tech Mathematics and Computing

Indian Institute of Technology, Delhi

AISSCE

St. Xavier’s Sr. Sec. School, Delhi

Interests

Large Language Models Computer Vision Reinforcement Learning Workflows
Publications
Car Drag Coefficient Prediction from 3D Point Clouds Using a Slice-Based Surrogate Model featured image
Deep Learning

Car Drag Coefficient Prediction from 3D Point Clouds Using a Slice-Based Surrogate Model

A lightweight slice-based model predicts vehicle drag (Cd) directly from 3D point clouds with high accuracy and fast inference.

avatar
Adarsh Roy
•
Read more
Projects
gthr featured image
CLI

gthr

gthr is a Rust CLI that lets you fuzzy-pick files or directories, then hit Ctrl-E to dump a syntax-highlighted Markdown digest straight to your clipboard and quit Saving to a file …

DE Models for Image Denoising featured image
Differential Equation Models

DE Models for Image Denoising

This project explores the use of non-local morphological Partial Differential Equations (PDEs) and p-Laplacian operators on graph structures for advanced image denoising and …

Experience

  1. Backend Engineer @ ML and Apps Platform

    Dragonfruit AI
    Worked across ML and Apps platforms. Optimized real-time pipelines and backend APIs, cut latency by >95%, and shipped production features in multimodal AI and enterprise apps.

  2. Backend Engineer

    Amazon
    Built fraud detection and payments features on AWS serverless stack, saving millions in logistics costs and streamlining EU operations.

  3. Volunteer Teaching Assistant (Linear Algebra & Differential Equations)

    Indian Institute of Technology, Delhi
    Assisted paper evaluations, problem classes, and student doubt sessions (Volunteer Role)

Education

  1. B.Tech Mathematics and Computing

    Indian Institute of Technology, Delhi
    GPA: 8.24/10.
    Relevant coursework: OS, Algorithms, Theory of Computation, Deep Learning, Data Mining, NLP.
    Read Thesis

  2. AISSCE

    St. Xavier’s Sr. Sec. School, Delhi
    Grade: 96.6%
Skills & Hobbies
Languages
Python, C/C++, Rust, Java, SQL, Bash, Lua
AI/ML
PyTorch, LangChain, Vector DBs (Qdrant), RAG, MCP, Scikit-learn
Cloud & Backend
AWS (S3, EC2, SNS, DynamoDB, SQS), Flask, GraphQL, REST, JWT
Redis, Celery, RabbitMQ, Grafana, ELK stack
Tools
GitHub Actions, CI/CD, Docker, Pre-commit hooks, W&B, Prompt Engineering
Technical Hobbies
Neovim
Terminal Workflows
Custom Ergonomic Keyboards/Mice
Typesetting (Typst)
Honors and Achievements
Best Student Paper AI-2025
BCS SGAI (AI-2025) ∙ August 2025
Awarded Best Student Paper at Cambridge University, work published in Springer Nature Lecture Notes in Computer Science (LNCS).
JEE Advanced 2020
IIT Joint Entrance Examination (Advanced) ∙ September 2020
All India Rank 285 among ~250k candidates.
JEE Main 2020
National Testing Agency (NTA) ∙ September 2020
All India Rank 570 among ~1.1M candidates.
BITSAT 2020
Birla Institute of Technology and Science (BITS) ∙ September 2020
Top 0.1% among ~300k candidates.
KVPY Scholar
DST, Government of India ∙ January 2020
All India Rank 445. Fellowship awarded by Government of India.
NSEA Merit Award
IAPT ∙ December 2019
National Top 1% (National Standard Examination in Astronomy).
Enactus National Competition — Winner
Enactus India ∙ August 2021
Led the winning presentation video, collaborated with a 50+ member team.
Languages
100%
English
100%
Hindi