Skip to content
Dhairya Mishra

My Story

Building systems that learn and scale.

I'm Dhairya, a Software Engineer and ML Researcher deeply fascinated by the intersection of distributed computing and artificial intelligence. My journey started with a simple curiosity about how machines understand the world, which eventually led me to NYU Courant to dive deep into AI.

Over the last 6+ years, I've had the privilege of building pipelines that serve millions at companies like CVS Health and Aetna. But beyond the sheer scale, what I truly care about is reliability and impact. I believe that the best ML architectures are those invisible to the user—quietly fast, undeniably robust, and gracefully degrading when the unexpected happens.

When I'm not orchestrating cloud-native backfills or fine-tuning language models, you can find me exploring new coffee shops, reading up on the latest multi-agent world models, or simply walking my dog while thinking through an architecture problem.

Engineering Philosophy

Principles that guide my day-to-day work.

01 Empathy for the User

Every millisecond of latency or skewed prediction is a degraded experience for someone. I build with an empathy-first mindset, ensuring the end-user feels the magic of AI without the friction.

02 Pragmatism over Hype

While I love experimenting with state-of-the-art models, production demands stability. I prefer robust, well-tested architectures that deliver consistent value over the newest, untested paradigm.

03 Observability is a Feature

If it isn't monitored, it isn't in production. I embed telemetry, structured logging, and robust alerting from day one, treating observability as a core requirement rather than an afterthought.

04 Continuous Learning

The AI landscape shifts weekly. I maintain a researcher's curiosity alongside an engineer's discipline, constantly reading papers, prototyping new ideas, and learning from past failures.

Publications & Research

I'm passionate about contributing back to the academic community. My core research interests lie in multi-agent reinforcement learning, video world models, and robust ML infrastructure.

Solaris: Multi-Agent Video World Models

ICML 2026 (Submission)

Researcher/Developer | NYU Courant | Sep 2025 – Present

Testaro: Web Accessibility Testing Framework

ACM SIGACCESS ASSETS 2023

Speaker/Contributor | CVS Health | Oct 2023

Academic Background

New York University

Courant Institute

M.S. Computer Science (AI)

May 2026 3.7

Trine University

B.S. Software Engineering & Mathematics

Dec 2021 Angola, IN