Arpan Narula

I like building products that are useful, fast, and easy to use, whether that means working on the frontend, the backend, or both.

Most of my projects sit somewhere between product, backend, and UI work. This portfolio is a mix of dashboards, APIs, support tools, legal tech, and a few smaller side projects.

View ProjectsGitHubResume
React + Next.jsNode + ExpressMongoDB + SQLAI Integrations

Quick intro

A mix of product, UI, and backend work

01
Product ideas
02
Frontend work
03
Backend logic
04
Deployment

Web apps

I enjoy building products that feel clean and practical.

APIs

A lot of my projects involve APIs, data flow, and integrations.

AI features

I like using AI where it adds something useful to the product.

Shipping

I care about actually finishing projects and putting them live.

Projects

6 projects on the site

Work I enjoy

Web apps, APIs, dashboards, and useful tools

Looking for

Software roles where I can keep building

Projects

Some things I’ve built

These are the projects I wanted to highlight first, with the smaller side work kept simple below.

Inventory Forecasting Platform

StockPilot AI

StockPilot AI helps businesses track inventory, spot deadstock, and make better restocking decisions using analytics and forecasting.

Inventory
Forecasting
Analytics

What I Worked On

  • Built APIs with Node.js and Express.js for inventory tracking, analytics, and forecasting flows.
  • Designed MongoDB models for users, products, SKU-level stock data, and forecast results.
  • Added processing pipelines for deadstock detection and restock recommendations.
  • Integrated Gemini to generate product insights and summaries inside the app.
  • Connected backend services with dashboards for real-time business visibility.

Tech Stack

Node.jsExpress.jsMongoDBPythonREST APIsGemini API
GitHubLive Demo

Quick Look

Dashboard
→
API Layer
→
Forecast Logic
→
MongoDB
Real-time inventory tracking
Forecasting and restock suggestions
Deadstock and SKU analytics
Business dashboards and insights

LLM Evaluation & Feedback Engine

EvalForge AI

EvalForge AI is a full-stack app that simulates how AI systems improve response quality using evaluation pipelines, ranking, retrieval, and feedback loops.

Evaluation
Ranking
Feedback

What I Worked On

  • Built a multi-stage pipeline that generates multiple responses, evaluates them, ranks them, and selects the best output.
  • Added scoring using rule-based logic, simulated embedding similarity, and simulated judge-style evaluation.
  • Implemented feedback handling so thumbs up and thumbs down can influence ranking weights over time.
  • Created retrieval logic to store stronger responses and reuse them for future prompts.
  • Split the app into a Next.js frontend and FastAPI backend with a clean full-stack flow.

Tech Stack

Next.jsReactFastAPIPythonRechartsVercelPytest
GitHubBackend APILive Demo

Quick Look

Prompt
→
Generate
→
Evaluate
→
Rank
→
Feedback
→
Improve
Multi-response generation
Evaluation and score breakdowns
Ranking and best-response selection
Feedback and retrieval loop

Current Build

Prototype system: generation and evaluation are simulated to show the architecture and workflow clearly.

Support Ticketing Platform

SupportOps Assistant

SupportOps Assistant is a support management tool for engineering teams to handle tickets, recurring issues, and debugging workflows in one place.

Support
Tickets
Analytics

What I Worked On

  • Built APIs for ticket creation, updates, status changes, severity, and ownership.
  • Designed MongoDB collections for ticket data, issue history, and performance metrics.
  • Added logic for spotting recurring issues and surfacing useful support insights.
  • Integrated external AI tools for root-cause suggestions during debugging.
  • Created analytics endpoints used by dashboards to track support performance.

Tech Stack

Node.jsExpress.jsMongoDBREST APIsClaude API
GitHubLive Demo

Quick Look

Support UI
→
Ticket APIs
→
Insight Logic
→
MongoDB
Ticket lifecycle management
Assignment and prioritization flows
Support analytics dashboards
AI-assisted debugging suggestions

Legal Research Platform

CourtIQ

CourtIQ is a legal research platform where users can upload documents, search through them, and get relevant case insights faster.

Documents
Search
Summaries

What I Worked On

  • Built APIs for document upload, processing, and search.
  • Handled metadata and indexing so legal content could be retrieved more efficiently.
  • Added summarization workflows to pull useful insights from large documents.
  • Connected the search interface with backend processing services and stored results.

Tech Stack

Node.jsExpress.jsFirebaseREST APIAI Workflows
GitHubLive Demo

Quick Look

Upload UI
→
Processing APIs
→
Search Logic
→
Firebase
Document upload and processing
Search across legal content
Document summaries
Faster research workflow

Personal Finance App

Finio

Finio is more of a UI and product design project, focused on budgeting flows, dashboards, and a clean finance experience for younger users.

React.jsNext.jsTailwind CSSFigma
GitHubLive Demo

Movie Review Sentiment App

IMDB Sentiment Analysis

A smaller ML side project that predicts sentiment for movie reviews and was useful for experimenting with NLP and model deployment.

PythonTensorFlowNLPHugging Face
GitHubLive Demo

Skills

Tech I use

Languages & Frameworks

Main stack

React.jsNext.jsNode.jsExpress.jsPythonTailwind CSS

Data & Backend

APIs and storage

MongoDBSQLFirebaseREST APIsSchema DesignAuthentication

Tools

Everyday workflow

GitGitHubVercelPostmanFramer MotionFigma
I’m comfortable working across the stack, depending on what the project needs.

Experience

Experience

Internship

Web Developer Intern

Innate Infotech

May 2025 - July 2025

Worked on both frontend and backend tasks, handled database changes, fixed bugs, and shipped product updates across ongoing client work.

Built and updated features using PHP and MySQL.
Handled bug fixes, UI improvements, and backend changes together.
Worked on day-to-day product updates in a fast-moving team setup.

Contact

Get in touch

Let's connect

Open to new opportunities

If you think I’d be a good fit for your team or project, feel free to reach out.

arpannarula9999@gmail.comGitHubLinkedIn