WAJIHA SYED

Full-Stack AI Developer specializing in LLM Integration and RAG Pipelines

Based in: Rawalpindi, PakistanMost recently: AI/ML DEVELOPMENT INTERN, Troon TechnologiesExperience: 2 years

Software Engineering graduate with extensive experience in building AI-driven applications using FastAPI, Python, and the MERN stack. Proven expertise in deploying containerized solutions that integrate OCR, NLP, and multimodal AI for complex technical challenges.

Experience

AI/ML DEVELOPMENT INTERN

Troon TechnologiesIslamabad, PakistaninternshipFeb 2025 – Sep 2026

Summary

Implemented backend APIs using Python, FastAPI, and MongoDB, enabling communication between application components and supporting structured data storage and retrieval. Integrated LLMs, RAG, OCR, and multimodal AI into application workflows, enabling AI capabilities to work alongside backend and frontend components. Implemented data-processing pipelines, validation logic, and API integrations, supporting consistent data handling and helping resolve application issues.

What I did

Focused on improving retrieval relevance, organizing the pipeline, and making backend workflows more efficient.
Automated the verification and approval/rejection process for building properties.
Balanced response latency with memory consumption when running LLMs locally on hardware with limited resources.

Results

Built a workflow that retrieves relevant context before generating answers to reduce reliance on the LLM's general knowledge.

PythonFastAPIMongoDBLLMsRAGOCRMultimodal AIPydantic

FRONTEND WEB DEVELOPMENT INTERN

Troon TechnologiesIslamabad, PakistaninternshipAug 2024 – Nov 2024

Implemented responsive web interfaces using React, JavaScript, HTML, CSS, and Bootstrap, translating application requirements into functional user-facing features. Built responsive layouts and frontend components, improving interface consistency across different screen sizes. Integrated frontend functionality with application workflows and debugged UI issues, helping deliver more reliable user interactions.

ReactJavaScriptHTMLCSSBootstrap

Skills

Technical

Python

Git

REST APIs

Object-Oriented Programming

GitHub

RAG

HTML

CSS

LLM Integration

FastAPI

NLP

MongoDB

Multimodal AI

AI Pipelines

pandas

NumPy

JavaScript

Docker

Containerized Deployment

Data Structures & Algorithms

Flask

Debugging

Node.js

OCR

React

Deep Learning

Bootstrap

Agent-Based Systems

Express.js

SCSS

Database Management

Data Storage and Retrieval

scikit-learn

Regression

Classification

Testing

Code Optimization

Oracle

Pydantic

Strict Schemas

Ollama

Local Model Deployment

Languages

Urdu

English

Projects

INTERNAL AI PRODUCT

Backend & AI Integrationprofessional

Implemented backend APIs using Python, FastAPI, and Node.js, enabling communication between application services and frontend components. Integrated MongoDB database workflows and data-processing logic, supporting structured data management across application features. Integrated OCR, LLM, and RAG components into AI-enabled workflows and used Docker for containerized deployment, making AI functionality available within the application environment.

PythonFastAPINode.jsReactDockerMongoDBREST APIsOCRLLMsRAGPydantic

AI-POWERED RESUME SCREENING SYSTEM

Full-Stack Developerpersonal

Implemented a resume-screening application using Python and Flask, providing a web-based workflow for processing candidate resumes. Integrated machine learning into the screening workflow, enabling automated resume evaluation and supporting candidate screening. Connected screening logic with Flask application endpoints, making resume evaluation functionality accessible through the application.

PythonFlaskMachine Learning

AI CHATBOT

Backend Developerpersonal

Implemented chatbot API endpoints using Python and Flask, enabling communication between the application and its users. Integrated MongoDB with the Flask backend, enabling persistent data storage and retrieval for database-driven workflows. Connected API logic with database operations, delivering a functional chatbot application with backend data management.

PythonFlaskMongoDBPydanticOllamaLlama

REAL-TIME WHITEBOARD WEB APPLICATION

Frontend Developerpersonal

Implemented an interactive browser-based canvas using HTML, CSS, and JavaScript, allowing users to create visual content through a web interface. Added JavaScript-based drawing interactions, translating user input into on-screen drawing actions. Integrated canvas functionality and user interactions into a single application, providing a browser-based workspace for visual collaboration.

HTMLCSSJavaScripthttps://github.com/Wajiha-Syed/whiteboard

MUSIC STREAMING WEBSITE

Full-Stack Developerpersonal

Implemented a React frontend connected to a Node.js and Express.js backend, linking user-facing functionality with server-side application logic. Integrated MongoDB and backend APIs, enabling data exchange between the frontend, backend, and database. Connected the MERN stack components into a full-stack web application, demonstrating end-to-end application development.

MongoDBExpress.jsReactNode.jsMERN Stack

NEUROSURG AI - AI-ASSISTED BRAIN TUMOR SEGMENTATION & SURGICAL PATH PLANNING

AI Developeracademic

Implemented deep-learning-based brain tumor segmentation on MRI scans, producing tumor-region predictions for medical image analysis. Combined tumor segmentation outputs with anatomical atlas information and no-go zones, incorporating anatomical constraints into surgical path-planning workflows. Applied the A* search algorithm to surgical path planning, generating candidate routes that account for defined anatomical constraints.

PythonDeep LearningMedical ImagingU-Net/nnU-NetA* SearchPydantic

Education

BS

International Islamic University Islamabad (IIUI)Software EngineeringSep 2026

Recognition

Certifications

Google Project Management

Google

Machine Learning Bootcamp

Not Specified

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

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

More people on Kanvis

Browse more people Create your own Kanvis