Profile chamber

Arya Mehta

Builder mode

Night-owl, architecture-first

Core strength

Turning complexity into systems

Signal

Backend + data + applied AI

About Arya

I build technical systems that can survive real use, not just real demos.

I'm Arya Mehta, a builder focused on backend engineering, data systems, and applied AI. I care most about work that takes messy inputs, sharp constraints, and real operational pressure, then turns them into something legible and reliable.

My strongest zone is where infrastructure, intelligence, and product execution overlap. That includes data platforms, retrieval systems, graph-aware workflows, and full-stack products where architecture quality matters as much as surface polish.

I like systems that are observable, scalable, and explainable to other engineers. I also like making them feel intentional, because strong engineering is not only about whether the system runs, but whether people can trust and extend it.

Capability lanes

The toolbox matters less than how the tools connect.

Data engineering

ETL/ELT pipeline designApache AirflowApache KafkaPySparkHadoopData lake designData quality

Cloud + scale

AWS EKS / S3Apache SparkDockerKubernetesCI/CD automationPerformance tuning

Applied AI

LangGraphLangChainRAG / Vector searchPyTorch / TensorFlowscikit-learnNLP & MLOps

Backend + product

Python / GoJava / Spring BootFastAPI / REST APIsReact / Next.jsWebSocketsSystem design

Proof markers

Signal that sits behind the introduction.

Hackathon execution

SJ Hacks winner

Built a strong product and technical story under pressure, then shipped it into a winning outcome.

Modeling discipline

Kaggle top placement

Proved strong experimentation instincts through competitive machine learning work and evaluation-driven iteration.

Engineering maturity

Open source contribution

Contributed back to tooling ecosystems instead of only consuming them, which signals deeper technical ownership.

Academic / event proof

Technical excellence recognition

Repeatedly stood out for building practical systems, not only for theory or presentation.

kernel_space.oshero0%