Arjun Adhikari

I am a Software Systems Engineer (Nepal Engineering Council Regd. No: 89754) with a background in Full-stack development, now focused on the systems engineering required to run AI models reliably at scale. My work centers on the belief that an AI model is only as good as the infrastructure supporting it.

My primary stack includes NestJS, Node.js, and BullMQ, which I use to manage distributed task queues and complex model orchestration. I applied this at scale for MintMyStory, a platform that manages multimodal content generation by coordinating multiple generative models. I also developed PharmaGuard, a blockchain-integrated tracking system that reached the semi-finals of a national innovation hackathon.

Key Projects & Achievements

• PillReminder (Edge AI): Developed an offline medication recognition app for rural clinics. Built a MobileNet V3 pipeline running via TensorFlow.js for entirely on-device inference. This project won Best Poster at the 1st National Conference on Patient Safety.

• AI Research: Published "The Illusion of Fairness" on TechRxiv, exploring unsupervised domain adaptation (UDA) on sparse data. I utilized Grad-CAM and EBM to interrogate and visualize model behavior.

• Engineering Impact: Currently contributing to the Plastic Bank project at Rumsan Associates, focusing on building transparent, impactful digital solutions.

Academic Foundation

I hold a Bachelor of Software Engineering from Pokhara University (3.80 CGPA, Dean's List). I thrive on solving the "boring" but critical problems of AI: latency, reliability, and edge-case orchestration.