What I Actually Do

Not a service menu. This is what I build.

01

AI & Machine Learning

I build neural networks that actually learn. Most of my work so far is in NLP, specifically emotion detection across multiple Indian languages, where direct translation loses the nuance the model needs. Training pipelines, evaluation, and getting a model out of a notebook and into something people can use.

Neural NetworksDeep LearningNLPModel TrainingModel Optimization
02

Full-Stack Product Development

I ship products end-to-end. AntarYatra went from an idea to a deployed platform I am still actively building on. I handle architecture decisions, performance trade-offs, and actually think about user experience. React to databases to deployment, I know the whole chain.

System DesignPerformanceScalabilityDevOpsUser Experience

Expertise Map

Skills ranked by depth and production experience

Ranked by how much I've actually built with each. Daily drivers are what I reach for on every project, and growing areas are where I'm actively leveling up.

Machine Learning

Daily Driver

Model training, deployment optimization, real-world validation

Backend Development

Daily Driver

APIs, databases, scalability, system design

Deep Learning

Deep Focus

Neural network architectures, NLP, training and evaluation

Frontend Development

Comfortable

React, TypeScript, UI/UX, animations

DevOps & Deployment

Comfortable

Docker, CI/CD pipelines, cloud infrastructure

ML Systems & Deployment

Growing

Data pipelines, inference optimization, serving models in production

The Tools I Use

Not just buzzwords. These are production tools I actually use.

Languages

PythonTypeScriptJavaScriptSQL

AI & ML

PyTorchHugging FaceScikit-learnNumPyPandas

Backend

Node.jsExpressFastAPIMongoDBRedisPostgreSQL

Frontend

ReactTypeScriptFramer MotionTailwind CSSNext.js

DevOps & Deployment

DockerVercelRailwayGitHub Actions

Looking to work together? Let's discuss your project.

Schedule a Call →