Moaz Nadeem

AI Engineer | LangGraph & RAG Specialist | Building Product-First AI for Scale

Based in: PakistanMost recently: Software Engineering Intern, Netroots Technology

Full-stack AI engineer specializing in agentic workflows and RAG systems. From architecting multi-agent research pipelines with LangGraph to building vision-augmented tutoring agents, I focus on shipping high-impact, production-ready AI. Back-to-back hackathon winner with a deep background in DevOps and scalable cloud architecture.

Experience

Software Engineering Intern

Netroots TechnologyinternshipJul 2026 – Sep 2026

Summary

Built an end-to-end e-commerce platform with an AI product comparison and recommendation engine, plus a chatbot for natural product filtering and discovery. Developed a complete ERP system for a shoe brand, covering core operational workflows and business data management.

What I did

Developed a chatbot for natural product filtering and discovery
Developed a complete ERP system for a shoe brand, covering 7 core operational workflows (inventory, sales, etc.) and business data management

Chatbot

Software Engineer Intern

Software Productivity StrategistsinternshipJul 2025 – Sep 2025

Summary

Redesigned a 15-page production website using React and Tailwind CSS, improving Lighthouse Performance by 98% and SEO by 10%. Implemented SEO, accessibility, and performance optimizations and collaborated with a 5-member team using Git branching and PR workflows.

Results

Improved Lighthouse Performance by 98%
Improved SEO by 10%

Tailwind CSSGit

Skills

Technical

JavaScript

TypeScript

Python

React

React Native (Expo)

Tailwind CSS

FastAPI

REST

LLM APIs (Groq, Tavily)

RAG

Supabase

Git

Node.js

Zustand

Next.js

PostgreSQL

NativeWind

pgvector

LangGraph

SSE Streaming

Supabase Auth (RLS)

Docker

SQL

AWS (S3, EC2, ECS)

Whisper (STT/TTS)

YOLOv8

CLIP

GitHub Actions (CI/CD)

SQLite

Bash

Terraform

Projects

GIKI Student Marketplace

Developer & DevOps Engineeracademic

Full-stack marketplace application for GIKI students with a containerized AWS deployment. Engineered a secure infrastructure using VPC, Security Groups, and IAM least-privilege roles. Automated the Docker build and push process to ECR using custom Bash scripts.

AWS EC2AWS S3AWS ECRDockerAWS VPCAWS IAMReactBash

ClipSeek - Natural-Language Video Search

Lead Developerpersonal

Built a full-stack video search application integrating YOLOv8, ByteTrack, and CLIP for natural-language retrieval and timestamped clip playback. Engineered a background indexing pipeline with frame sampling, appearance deduplication, and batched embeddings; remote GPU execution ran 4.27× faster than local (22.6s vs 96.2s) on a sample video, excluding upload and database I/O.

ReactTypeScriptFastAPIPostgreSQLYOLOv8ByteTrackCLIP

Visit the site View the code

Company Research Agent - Multi Agent Research Pipeline

Architect & Developerpersonal

Architected a 4-agent LangGraph workflow (Clarity, Research, Validator, Synthesis) with conditional routing and human-in-the-loop clarification, integrating Tavily search and an LLM with deterministic confidence scoring (0-10) and up to 3 research retry loops. Built a REST API with FastAPI including SSE streaming endpoints, with persistent state via LangGraph checkpoints (Postgres fallbacks, Supabase integration) for thread-based conversations.

LangGraphFastAPITavilyLLM APIsPostgresSupabaseSSE

View the code

Uraan - RAG-Based Voice AI Tutor for Students

Lead Developerpersonal

Built an end-to-end voice tutoring agent: speech-to-text → textbook-grounded LLM answer → text-to-speech, with Whisper fallback for STT reliability. Designed a RAG pipeline over FSc Biology textbooks using local/remote embeddings (nomic-embed-text) and Supabase pgvector cosine retrieval with page-level citations. Implemented vision-augmented ingest: PDF/OCR extraction + VLMs to caption figures/tables, stored as typed chunks (text / figure / table) for diagram-aware answers.

FastAPINext.jsGroqSupabasepgvectorDockerWhispernomic-embed-textOCRVLM

Visit the site View the code

Education

Bachelor of Science

Ghulam Ishaq Khan Institute of Engineering Sciences & TechnologyComputer ScienceSep 2023 – May 2027

Coursework

Software EngineeringDatabasesDevOpsComputer NetworksOperating SystemsOOPDSAAlgorithm AnalysisMachine Learning

Recognition

Awards

3rd Place | AI Hackathon - Hack at GIKI (2026)

GIKI2026

Competed against 50+ participants

2nd Place | ICPC Regionals Onsite (2025)

ICPC2025

Ranked 2nd among 20+ GIKI teams and 21st among 100+ teams regionally

Runner-Up | Agentic AI Hackathon (2025)

GDGOC GIKI2025

1st Place | Skillo AI Developer Challenge

Skillo

Ranked 1st among 400+ participants

2nd Place | Comeback Pakistan

Comeback Pakistan

Ranked 2nd among 15 teams and 60+ builders (built Uraan)

Governance & advisory

Development and Marketing

Microsoft Club GIKIother2024 – Present

Developed university software i.e. ORIC Dashboard for 100+ professors. Co-head end-to-end event delivery for a 36-hour inter-university hackathon including planning and sponsor acquisition. Led promotional efforts with a team of 10.

What drives the work

Motivations: AI Agents
RAG Systems
Full-Stack AI
Computer Vision
Early-Stage Startups
Product Engineering

Hire MoazReach out about a role, a contract or a conversation.For recruiters

Moaz's twin is AI, it can make mistakes.

More people on Kanvis

Browse more people Create your own Kanvis