Resume Contact
About Achievements Experience Projects Gallery ReadList
Naman Agarwal. SOFTWARE DEVELOPER
Web
Android
Backend
[naman@root]$
whoami
I am a passionate Full-Stack Developer currently working as a Software Development Engineer 1 (SDE 2 in-line) at Juspay, a company that builds robust and scalable rails for digital payments. I specialize in building reliable and scalable solutions for real-world problems. My experience spans functional systems programming, microservices, and Generative AI pipelines. I take a bottom-up approach to learning—I strive to understand tools from first principles, often reverse-engineering or building them from scratch (e.g., AI Agents, React, JSON parsers) to truly master them. I thrive on challenges and have a proven ability to make critical engineering decisions to deliver high-quality, high-impact projects.
[naman@root]$
cat ./tools.json
Achievements
Awarded 'Survival Startup' title at the UBC Hackathon
Conducted research in privacy-preserving Federated Machine Learning applications
Awarded a Letter of Recommendation from the Additional Director General of Uttar Pradesh, India.
Served as Tech Lead for the Android Club at VIT Chennai
Architected and deployed a full-stack productivity application to the Google Play Store, managing end-to-end delivery
[naman@root]$
ls ./experience
Juspay
SDE 1
  • Trident Server: Led the Rust port of a heavily utilized 3DS server, currently live with major enterprise clients like HDFC and ixigo. The new implementation increased overall transaction success rates by 2-3%.
  • Godel AI: Brainstormed, architected, and brought to production an AI-driven system to autonomously maintain and run scrapers for bank OTP pages. This full-stack solution drastically reduced operational overhead for merchants like Swiggy, Zepto, and Myntra, requiring only human-in-the-loop approvals.
  • Agentic Payments: Developed and integrated coding agents for product discovery and payment workflows utilizing the Model Context Protocol (MCP).
  • Infrastructure & DevOps: Hosted and maintained multiple critical services utilizing Docker.
  • Technologies: Rust, Haskell, Python, JavaScript, Docker, LLMs, Prompt Engineering, Clean Architecture.
ZeeQ
  • Architected highly scalable back-end services from the ground up utilizing Clean Architecture principles and Docker.
  • Developed over 40% of the back-end services, including a robust authentication and authorization system with role-based access control.
  • Translated business requirements into technical specifications and managed the database with Hasura ORM.
  • Technologies: Docker, Node.js, GraphQL, Git.
Pharmiq
  • Engineered robust back-end systems in Go and Python, implementing secure authentication middleware and optimizing state management.
  • Technologies: Docker, Go, Python, TypeScript, Git, Jira.
GRP India
  • Delivered a desktop application to automate duty allotment to station personnel, reducing allotment time by 50%.
Zoho
  • Collaborated with the machine learning research team to develop and evaluate privacy-preserving Federated Learning models utilizing TensorFlow and Python.
  • Technologies: TensorFlow, Python.
[naman@root]$
less projects.list
Text-to-Video Pipeline
Architected and built a fully autonomous Text-to-Video pipeline where AI agents generate complete videos from a single text prompt, eliminating the need for manual video editing.
  • Engineered AI-driven content analysis utilizing LLMs, precise prompt engineering, and embeddings to ensure high thematic relevancy between the story and generated video clips.
  • Designed and hosted the rendering pipeline on AWS Lambda to efficiently handle low-traffic, high-memory, and bursty computational requirements, demonstrating a strong understanding of cloud pricing and architectural tradeoffs.
  • Implemented AI models to autonomously maintain visual and narrative consistency across all video segments.
  • Empowered the AI to independently search and integrate real-time data and media via APIs.
Family Tree Builder (Computer Vision / Data Eng)
Engineered a data processing pipeline to construct complex relational graphs from public Indian electoral data. Developed custom Computer Vision tools integrating Tesseract OCR to read image-based PDFs, built automated CAPTCHA solvers, and scraped/normalized 25,000+ data points to map multi-generational relationships.
Invoice Central
An invoice-generating web app for offline shops, featuring product listing, quantity management, and invoice creation/sharing. Optimized database queries with MongoDB's Aggregation pipeline, improving response times by 5%.
Confab Video Chat
Built a video chat web app using WebRTC and WebSockets for full-duplex communication. Implemented efficient tree-based data structures to increase multi-user video chat performance by 10%.
Face Recognition application
Developed an automated attendance system utilizing Google's Facenet deep learning model and MQTT for IoT integration.
React Native Transition View
Authored a highly performant open-source library leveraging Reanimated 2 to handle complex transition calculations efficiently at 60fps.
[naman@root]$
cat ./contact.js
Email
namanddev7@gmail.com
Phone
9651111184
LinkedIn
Github
Social