Ali Irfan

AI / Python Backend Engineer specializing in LLM & Agentic AI Applications

Based in: Lahore, PakistanCurrently: Backend & Full-Stack Developer (Freelance), UpworkExperience: 1 year

Hire me

Computer Science graduate with expertise in building AI-powered backend systems using Python, FastAPI, and Django. Experienced in developing end-to-end multimodal AI systems, agentic customer-support applications, and fine-tuning LLaMA-3 models.

Experience

Backend & Full-Stack Developer (Freelance)

Upworkfreelance2026 – Present
Job Success Score
100%Job Success Score
rating with 7 client reviews
5.0rating with 7 client reviews

Summary

Completed 12 fixed-price projects with 100% Job Success and a 5.0 rating (7 client reviews); delivered web, mobile and AI-based solutions, gathering requirements and shipping production-ready applications. Worked with Python (FastAPI / Django) and databases, and troubleshot issues in client projects.

What I did

  • Developed an AI-based plant recommendation system using Django, FastAPI, and machine learning models to predict plant behavior.

Results

  • Successfully delivered 12 fixed-price projects
PythonFastAPIDjangoDatabasesMachine Learning

Python Backend Intern

Dexterz Technologies (Pvt.) Ltd.LahoreinternshipJul 2025 – Oct 2025
Developed backend modules in Python with the engineering team using Python frameworks and relational databases. Followed professional coding standards, Git-based collaboration and backend optimization practices.
PythonRelational DatabasesGit

Skills

Technical

PythonPython
DjangoDjango
FastAPIFastAPI
Django REST FrameworkDjango REST Framework
GitHubGitHub
PostgreSQLPostgreSQL
GitGit
LangChain
SQLSQL
REST API design
LangGraph
RAG
AI Agents
Deep Learning
LLaMA-3 fine-tuning
YOLO
C++C++
SQLAlchemySQLAlchemy
Machine Learning
NLP
Computer Vision
MySQLMySQL
HTMLHTML
CSSCSS
BootstrapBootstrap
Web Scraping
Automation
ReactReact
FirebaseFirebase
Model Quantization
Java
Pydantic
Gemini Flash 2.0

Projects

MedPercept AI - Vision Language Patient Intent System (Final Year Project)

Lead Developeracademic
Built a real-time multimodal AI system that combines four parallel computer vision models with a fine-tuned LLaMA-3 reasoning model to analyze patient behavior and generate natural-language intent analysis and alerts. Developed the Django backend, REST APIs, SQL database integration and frontend components, connecting the full AI pipeline (vision → reasoning → alerts) into one working application. Optimized the models for edge deployment using quantization and efficient fine-tuning techniques.
PythonYOLOLLaMA-3DjangoREST APIsSQLQuantizationFine-tuningLLaMA reasoning modelPostgreSQL

My Gardener - AI-Based Plant Care System

Backend Developerpersonal
Built a web app for plant tracking, care reminders and AI-based recommendations by plant type. Developed backend APIs for plant and user management using Django, FastAPI and PostgreSQL.
DjangoFastAPIPostgreSQLHTMLCSSBootstrap

Tigrinya Learning App

Mobile Developerpersonal
Designed, built and published a mobile language-learning app on the Google Play Store, covering the user interface and learning features.
Mobile App Development

Voice-to-Text Emotion Detection System

AI Developerpersonal
Built a speech-to-text pipeline that classifies emotion (positive, negative, neutral) from voice input using NLP techniques.
PythonNLPSpeech-to-Text

ShopAgent - AI Customer Support Agent

Full-Stack Developerpersonal
Built a full-stack agentic AI customer-support system with a LangGraph workflow, RAG-based answers, tool calling and short-term conversation memory. Added Human-in-the-Loop approval for sensitive actions; exposed the agent through a FastAPI backend with PostgreSQL storage and a React interface.
LangGraphLangChainRAGFastAPIReactPostgreSQLTool CallingHuman-in-the-LoopPydantic

Education

BS Computer Science

University of Central PunjabComputer ScienceGPA 3.05Oct 2022 – Jul 2026

Thesis. MedPercept AI - Vision Language Patient Intent System

Coursework

Backend developmentAI applicationsSoftware engineering
Hire AliReach out about a role, a contract or a conversation.For recruiters

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