Skip to content

Profile / AG-01

Available

GenAIFull Stack Engineer

AdityaGupta

Full-stack engineer shipping production React, Python and AWS products end to end — including dependable LLM features inside real user workflows.

View selected workEmail me
GitHub ↗LinkedIn ↗
Next sectionWhat I own

One person owns all four layers.

frontend

Interfaces people actually use

React and Next.js frontends serving thousands of users at high request volume, taken from spec to production release. Responsive and load-efficient, with usability across devices as the constraint rather than an afterthought.

  • React
  • Next.js
  • Redux
  • TailwindCSS

backend

The services behind them

Python backends and REST API design over MongoDB, with authentication workflows and Stripe subscription lifecycle handling — including the secure transaction paths that come with taking money.

  • Python
  • REST
  • MongoDB
  • Stripe

infra

The ground it runs on

Production deployments on AWS EC2 with Docker and Nginx, managed for high availability, scalability, and reliable rollouts. CI/CD through GitHub Actions.

  • AWS EC2
  • Docker
  • Nginx
  • CI/CD

ai.layer

Model output, made dependable

LLM-backed features wired into live application flows: prompt design, working with structured model outputs, and handling the failure cases so what reaches the user stays reliable.

  • LLM APIs
  • Prompt design
  • Structured outputs
  • NLP

AI work

A model call is the easy part.

Wiring an LLM into a live product is mostly everything around the call: composing a prompt from real application context, getting structured data back instead of prose, and deciding what the user sees when the model is slow, wrong, or unavailable.

POST /feature/generate

1,284 ms

The slow, unreliable step. Everything around it exists to keep that from being the user's problem.

cost
dominates the request
assume
it can fail

Illustrative flow — the shape of the work, not a capture from a production system. Durations are indicative.

Timeline

2021 — now

Eighteen months, drawn to scale.

Real dates, drawn to scale from the first engineering job. The degree runs off the left edge — it was still going when the first one started.

Jul 25
Nov 25
Mar 26
Jul 26

Pranveer Singh Institute of Technology

B.Tech · Computer Science (AI & ML)

3y 8m

Wisprout Life Pvt. Ltd.

Software Engineering Intern

6m

Allcognix AI

GenAI Full Stack Developer, Intern

6m

Allcognix AI

GenAI Full Stack Developer

ongoing

| marks the promotion from intern to engineer, six months in.

Experience

3 roles

The record, in full.

Three roles across two companies, and a promotion six months in. Nothing here is hidden behind an interaction.

Allcognix AI

GenAI Full Stack Developer

Mar 2026 — Presentongoing

Promoted from GenAI Full Stack Developer Intern after six months.

  • Lead full-stack development across React/Next.js frontends and Python backends serving thousands of users at high request volume, taking features from spec to production release.
  • Own production deployments on AWS EC2 with Docker, managing environments for high availability, scalability, and reliable rollouts.
  • Own Stripe payment integration, covering subscription lifecycle handling and secure transaction workflows.
  • Integrate LLM-backed features into the product, wiring model responses into application flows and handling failure cases so results stay reliable for users.
  • Build responsive, load-efficient interfaces with a focus on usability across devices.
  • Partner with product and engineering peers to align frontend, backend, and AI systems into one coherent product.
  • React
  • Next.js
  • Python
  • AWS
  • Docker
  • Stripe
  • LLM APIs

Allcognix AI

GenAI Full Stack Developer, Intern

Sept 2025 — Mar 20266m

  • Developed and integrated AI/ML features in Python, wiring model outputs into live application flows.
  • Built responsive UI components and connected frontend–backend API communication workflows.
  • Supported Docker-based containerization, AWS EC2 setup, and Stripe payment gateway implementation.
  • Python
  • React
  • Docker
  • AWS EC2
  • Stripe

Wisprout Life Pvt. Ltd.

Software Engineering Intern

Mar 2025 — Sept 20256m

  • Built features for a secure, scalable coaching platform supporting client self-enrollment, coach selection, and session scheduling.
  • Contributed to AI-generated session notes and summaries, converting raw session content into structured, client-facing output.
  • Implemented outcome analytics benchmarking client progress against internal and external market data to quantify coaching effectiveness.
  • Full stack
  • AI summarisation
  • Analytics

Projects

2 builds

Built outside the job, for the same reasons.

Two things I shipped end to end — a classifier with a dashboard in front of it, and a search interface over a headless catalog.

Twitter Sentiment Analysis

May 2025

Source

An NLP classifier that sorts tweets into positive, negative and neutral at 85%+ accuracy, paired with a Next.js dashboard for exploring the results.

collect · tweepy

preprocess · text pipeline

train · scikit-learn

serve · next.js dashboard

  • Collected live Twitter data via Tweepy and built the text preprocessing pipeline.
  • Trained and evaluated classifiers in Scikit-Learn, reaching 85%+ accuracy.
  • Built a Next.js dashboard for exploring classified results.
  • Python
  • NLP
  • Scikit-Learn
  • MongoDB
  • Next.js
  • TailwindCSS

Trendsetter

Dec 2024

Source

A responsive product-discovery interface with keyword search running across a headless CMS catalog.

fetch · wix headless

search · keyword index

render · responsive grid

  • Built keyword search across a headless CMS product catalog.
  • Optimised layout and rendering for a consistent experience across mobile and desktop.
  • JavaScript
  • React
  • Next.js
  • Wix Headless

Stack

32 entries

Everything on the panel.

Frontend

  • ReactJS
  • NextJS
  • Redux
  • TailwindCSS
  • Responsive UI
  • Performance-focused UI

Backend & Data

  • Python
  • REST API design
  • MongoDB
  • Stripe subscriptions
  • Stripe payments
  • Authentication workflows

DevOps & Cloud

  • AWS EC2
  • Docker
  • Nginx
  • Git
  • GitHub Actions
  • CI/CD
  • Production environments

AI & ML

  • LLM API integration
  • Prompt design
  • Structured model outputs
  • NLP
  • Scikit-Learn

Languages

  • Python
  • JavaScript
  • SQL
  • C++

Other

  • Figma
  • Blender
  • English (fluent)
  • Hindi (fluent)

Record

  • B.Tech, Computer Science (AI & ML)Pranveer Singh Institute of Technology (AKTU), KanpurNov 2021 – Jul 2025
  • Class XII — 86% · Class X — 83%Delhi Public School, Kanpur
  • PythonHackerRankCertification
  • C++HackerRankCertification
  • Full Stack DevelopmentUdemyCertification

Automotive systems

Off the clock · still technical

Cars are systems you can feel.

I am interested in more than the badge or the headline number. The interesting part is how input becomes response — through sensors, control logic, mechanical systems and the interface in front of the driver.

Reference car

RB19 · 2023 regulations

Side elevation · WebGL unavailable

Vehicle signal path

Input → response → feedback

  1. 01

    Input

    Sensors · driver intent

    Everything starts with a clean signal and the context around it.

  2. 02

    Process

    ECU · control logic

    Fast decisions, constrained by the system they have to protect.

  3. 03

    Output

    Powertrain · chassis

    The result has to be predictable, responsive and usable.

  4. 04

    Feedback

    Telemetry · driver feel

    Observe what happened, diagnose it, then make the next pass better.

Systems I keep reading about

  • Powertrain

    torque delivery

    How combustion, boost, gearing and electrification shape the response at the pedal.

  • Vehicle control

    sensors → decisions

    The conversation between sensors, ECUs and the mechanical systems under them.

  • Driver interface

    information at speed

    Clusters, warnings and controls that deliver the right information without demanding attention.

  • Race engineering

    performance over a stint

    Tyres, energy, aero balance and strategy treated as one connected system.

Race-weekend reading

Formula 1

Max Verstappen · Oracle Red Bull Racing

  • I follow the engineering behind the lap: tyre windows, energy deployment, aero balance and the strategy that connects them.
  • The part I enjoy most is the feedback loop — measure, interpret, adjust, repeat.

Same engineering instinct

Observe. Diagnose. Tune. Validate.

Whether the system is running on a server or moving on four wheels, I want to know what the signals say and why it behaves the way it does.

Open to remote & relocation.

If you are hiring for full-stack product work — or for someone who can own the AI features alongside it — the fastest way through is email.

adityaoff.1007@gmail.com

Kanpur, India · Open to remote & relocation