Seattle, WA

I build software systems powered by intelligence.

I design AI agents, orchestrated workflows, and intelligent automations that turn intent into dependable outcomes.

Explore my work Read my publication

I'm pursuing my M.S. in Computer Science at Georgia Tech, specializing in Artificial Intelligence.

Human in the loopDjay / orchestrator

System online · select a node to inspect the workflow.

NowSoftware EngineerAmazon · Seattle

StudyM.S. Computer ScienceGeorgia Tech · AI

Impact1,000+ usersAcross products shipped

ResearchPublished co-authorMicroscopy & Microanalysis

43.0731° N / 89.4012° W

01 / About

I'm Dhananjayan
Djay to most people.

I'm an AI engineer based in Seattle. I started by building apps, data products, and cloud systems; now I design AI agents, intelligent workflows, and the software that makes them dependable.

I'm pursuing my M.S. in Computer Science at Georgia Tech, specializing in Artificial Intelligence. I care about what happens after the demo: context, tool design, evals, observability, and the human judgment that turns a capable model into a useful system.

NowM.S. Computer ScienceGeorgia Tech · Artificial Intelligence

2024—NowSoftware EngineerAmazon · Merch on Demand

2022—2024Software DeveloperUW–Madison Cryo-EM Research Center

2020—2024B.S. Computer Science & Data ScienceUniversity of Wisconsin–Madison

View my résumé Connect with me

02 / Current focus

From writing code to orchestrating outcomes.

I'm interested in the software beyond a single model or chat box: designed systems of context, tools, agents, feedback loops, and human judgment.

01 Active thread

AI agent systems

I design bounded, tool-using agents that can plan, take action, inspect reality, and recover when the path changes.

02 Active thread

Intelligent workflows

I combine deterministic automation with adaptive reasoning so recurring work becomes faster, clearer, and more resilient.

03 Active thread

AI-native engineering

I'm reworking the development loop around clear intent, capable agents, measurable evals, and ownership of the outcome.

AI lab / building in public

I'm building new agentic case studies now. I'll share them when there is real work to show—clear architecture, honest constraints, and evidence that the system performs.

03 / Published research

Research that turned complex data into usable workflows.

Before the current shift to agentic systems, I helped build scientific software for high-throughput cryo-electron tomography—where orchestration, observability, and reliable pipelines were already essential.

Journal article Microscopy & Microanalysis · Vol. 29

TomoFlows: Pre-Processing Workflows for Cryo-Electron Tomography

Matthew R. Larson, Yan Zhuang, Djay Pallavur, Jae Yang, Bryan Sibert, Elizabeth R. Wright

I co-authored this work on a modular processing pipeline that wraps motion correction, IMOD commands, and machine-learning approaches for GPU-aided computing—helping microscopy facilities evaluate incoming data and generate preview tomograms faster.

Read my paper Explore TomoFlows

DOI · 10.1093/micmic/ozad067.357

04 / Shipped work

The builds that formed my foundation.

Products, experiments, and hackathon builds from my software-development chapter. These are the systems that taught me to ship across web, mobile, cloud, data, and machine learning.