Umair Amjad Khan — AI & Machine Learning Engineer | MLOps & LLMOps Specialist

AI & Machine Learning Engineer

Umair Amjad Khan is an AI and Machine Learning Engineer and the Founder & CEO of icode Studios. He specializes in building production-grade AI systems, Large Language Model (LLM) pipelines, Retrieval Augmented Generation (RAG) architectures, Computer Vision solutions, MLOps workflows, and LLMOps pipelines. Based in Pakistan, Umair works with clients worldwide delivering intelligent software applications.

Umair Amjad Khan — AI and Machine Learning Engineer
icode StudiosAI Engineering & Custom Software Development Studio — you imagine, WE code

About Umair Amjad Khan

Umair Amjad Khan is the Founder and CEO at icode Studios, an AI engineering and custom software development studio. He is an AI and Machine Learning Engineer specializing in building production-grade intelligent systems. At icode Studios, he leads software architecture and AI engineering initiatives, delivering custom RAG pipelines, LLM fine-tuning solutions, and scalable web and mobile applications for clients worldwide. His technical expertise spans deep learning, computer vision, and full-stack MLOps — containerizing models with Docker, exposing high-throughput FastAPI backends, and building interactive software products.

MLOps & LLMOps Expertise of Umair Amjad Khan

Umair Amjad Khan is an MLOps and LLMOps specialist who builds end-to-end machine learning operations pipelines for production AI systems. His MLOps expertise includes Docker containerization, CI/CD pipelines for ML model deployment, model monitoring and observability, feature stores, and model registries. His LLMOps expertise includes LLM evaluation frameworks, guardrails for LLM safety, prompt versioning systems, LLM observability with LangSmith, and production deployment of Large Language Model pipelines. Umair Amjad Khan designs and implements MLOps and LLMOps workflows that ensure AI systems run reliably at scale in production environments.

Skills & Technologies of Umair Amjad Khan

  • Python
  • PyTorch
  • TensorFlow
  • Machine Learning
  • Deep Learning
  • Large Language Models (LLMs)
  • LLM Fine-Tuning (LoRA, QLoRA)
  • Retrieval Augmented Generation (RAG)
  • Computer Vision
  • Object Detection (YOLO)
  • Natural Language Processing (NLP)
  • Multi-Agent Systems
  • Agentic Workflows
  • Reinforcement Learning
  • Generative AI
  • MLOps
  • LLMOps
  • ML Pipeline Orchestration
  • Model Monitoring & Observability
  • LLM Evaluation & Guardrails
  • LLM Observability (LangSmith)
  • Prompt Versioning
  • CI/CD for ML
  • Docker
  • FastAPI
  • LangChain
  • LangGraph
  • LlamaIndex
  • OpenCV
  • Prompt Engineering
  • Vector Databases
  • Neural Networks
  • Transformer Architecture
  • Next.js
  • React
  • TypeScript
  • Flask
  • Git & GitHub
  • REST APIs

Featured AI Projects by Umair Amjad Khan

Zari.AI — Agricultural Disease Detection

Umair Amjad Khan built Zari.AI, a voice-enabled AI system providing real-time crop disease diagnosis via web and WhatsApp interfaces. Zari.AI helps farmers in rural areas detect diseases and get actionable treatment recommendations instantly through natural conversation. Built with Python, Computer Vision, Speech Processing, FastAPI, WhatsApp API, and PyTorch. Features real-time disease detection via camera input, multi-language voice interface, WhatsApp integration for field accessibility, and FastAPI backend with sub-200ms response.

Deepfake Detection Platform — ResNeXt + LSTM Pipeline

Umair Amjad Khan developed an end-to-end deepfake video verification pipeline combining ResNeXt spatial feature extraction with LSTM temporal analysis. Includes a Flask backend API and a Chrome Extension for real-time browser-based verification of media authenticity. Built with PyTorch, ResNeXt, LSTM, Flask, Chrome Extension API, and OpenCV. Achieves 90%+ detection accuracy on benchmark datasets with frame-level confidence scoring.

Autonomous Cyber Defense System — Multi-Agent Reinforcement Learning

Umair Amjad Khan created a hierarchical multi-agent reinforcement learning pipeline for real-time network threat detection and automated mitigation. Agents autonomously identify intrusion patterns, classify attack types, and deploy countermeasures without human intervention. Built with Python, PyTorch, Multi-Agent RL, Network Analytics, and Gymnasium.

Experience & Career of Umair Amjad Khan

Founder & CEO — icode Studios

2025 — Present | Pakistan & Remote

Umair Amjad Khan founded and leads icode Studios — an AI engineering and custom software development studio delivering production AI systems, LLM pipelines, and full-stack applications. Leading AI product strategy, system architecture, and client solution design. Building custom RAG architectures, LLM fine-tuning, agentic workflows, and end-to-end MLOps and LLMOps pipelines.

Machine Learning Engineer Intern — Avant Labs

2024 — Present | Islamabad, Pakistan

Umair Amjad Khan developed and trained Machine Learning models and Convolutional Neural Networks (CNNs), with hands-on experience in Flutter mobile app integration and backend pipeline optimization.

Freelance AI Engineer — Upwork & Fiverr

2023 — Present | Remote — Worldwide

Umair Amjad Khan delivers end-to-end AI and computer vision solutions for international clients. Built custom object detection and NLP pipelines, AI-powered chatbots, and intelligent backends.

BS Artificial Intelligence — University of Haripur

2022 — 2026 | Haripur, Pakistan

Umair Amjad Khan completed a Bachelor of Science in Artificial Intelligence focusing on Computer Vision, Deep Learning, and real-world AI applications.

Incoming MS Artificial Intelligence — GIKI

2026 | Topi, Swabi, Pakistan

Umair Amjad Khan is pursuing a Master of Science in Artificial Intelligence at GIKI — Ghulam Ishaq Khan Institute, one of Pakistan's premier engineering institutions.

Advanced AI Bootcamp 2026 — GIKI

2026 | Topi, Swabi, Pakistan

Umair Amjad Khan attended an intensive Advanced AI Bootcamp at GIKI covering Generative AI, RAG, Large Language Models, production deployment pipelines, and end-to-end MLOps & LLMOps pipeline engineering.

Certifications of Umair Amjad Khan

  • Certificate of Completion: Claude 101 — Anthropic (Jul 2026)
  • AI Fluency Framework & Foundations — Anthropic (Jul 2026)
  • Develop Generative AI Applications: Get Started — IBM (Jun 2025)
  • Build RAG Applications: Get Started — IBM (Jun 2025)
  • Generative AI and Model Selection — Vanderbilt University (Feb 2025)
  • AI For Everyone — DeepLearning.AI (Feb 2025)
  • AI Agents in LangGraph — Coursera / DeepLearning.AI (2026)
  • Python for Data Science, AI & Development — IBM / Coursera (2026)
  • Foundations of Cybersecurity — Google / Coursera (2026)
  • Hack N' Connect — GIKI (Sep 2023)

Frequently Asked Questions about Umair Amjad Khan

Who is Umair Amjad Khan?
Umair Amjad Khan is an AI & Machine Learning Engineer and the Founder & CEO of icode Studios. He specializes in building production-grade AI systems, LLM pipelines, RAG architectures, Computer Vision solutions, MLOps, and LLMOps. He holds a BS in Artificial Intelligence from the University of Haripur and is pursuing his MS AI at GIKI.
What is Zari.AI?
Zari.AI is a voice-enabled agricultural AI system built by Umair Amjad Khan. It provides real-time crop disease diagnosis via web and WhatsApp interfaces, helping farmers in rural areas detect diseases and receive actionable treatment recommendations through natural conversation.
What are Umair Amjad Khan's core skills in AI?
Umair Amjad Khan's core skills include Machine Learning, Deep Learning, Large Language Models (LLMs), LLM Fine-Tuning (LoRA/QLoRA), Retrieval Augmented Generation (RAG), Computer Vision, MLOps, LLMOps, Multi-Agent Systems, Agentic Workflows, Natural Language Processing, and Generative AI. He is proficient in Python, PyTorch, TensorFlow, LangChain, LangGraph, Docker, and FastAPI.
What is icode Studios?
icode Studios is an AI engineering and custom software development studio founded by Umair Amjad Khan. The studio's motto is "you imagine, WE code." It delivers production AI systems, LLM pipelines, RAG architectures, and full-stack web and mobile applications for clients worldwide.
What MLOps and LLMOps tools does Umair Amjad Khan use?
Umair Amjad Khan uses Docker for containerization, FastAPI for high-throughput API backends, CI/CD pipelines for ML model deployment, model monitoring and observability tools, LLM evaluation frameworks, guardrails for LLM safety, LangSmith for LLM observability, and prompt versioning systems. He builds end-to-end MLOps and LLMOps pipelines for production AI systems.
Who is the AI Engineer behind Zari.AI?
The AI Engineer behind Zari.AI is Umair Amjad Khan, an AI & Machine Learning Engineer and the Founder & CEO of icode Studios. He built Zari.AI as a voice-enabled agricultural disease detection system to help farmers in rural areas of Pakistan.

Contact Umair Amjad Khan

Email: umairamjadkhanamazai@gmail.com

WhatsApp: +92 317 0478541

Location: Pakistan — Available Remotely Worldwide

GitHub: github.com/Uak69009

LinkedIn: Umair Amjad Khan on LinkedIn

Fiverr: umair_khan69009 on Fiverr

Welcome to my digital workspace.

I'm Umair Amjad Khan—a Machine Learning Engineer and AI Specialist dedicated to building production-grade AI systems, LLM pipelines, and intelligent software applications. Whether you're exploring my latest open-source research, reviewing deployment architectures, or looking to collaborate on custom AI solutions, I'm glad you're here.

Explore my featured projects below, or feel free to reach out directly to start a conversation.

Scroll down to explore
Umair Amjad Khan — AI and Machine Learning Engineer
Available for Opportunities

Umair
Amjad Khan

Founder & CEO @ icode StudiosAI & ML Engineer

Spearheading icode Studios — Building Production AI Systems, LLM Pipelines, and Intelligent Applications that solve real-world problems.

10+
Projects Built
5+
AI Models Deployed
21
GitHub Repos
About Me

Turning Data into Intelligent Systems

I'm Umair Amjad Khan, Founder & CEO at icode Studios (“you imagine, WE code”) and an AI & Machine Learning Engineer specializing in building production-grade intelligent systems.

At icode Studios, I lead software architecture and AI engineering initiatives, delivering custom RAG pipelines, LLM fine-tuning solutions, and scalable web & mobile applications for clients worldwide.

My technical expertise spans deep learning, computer vision, and full-stack MLOps & LLMOps — containerizing models with Docker, building CI/CD pipelines for ML, deploying LLM evaluation guardrails, exposing high-throughput FastAPI backends, and building interactive software products.

LLM Fine-TuningRAG ArchitecturesComputer VisionMLOps PipelinesLLMOps PipelinesMulti-Agent Systems
3+
Years of Experience
10+
Projects Delivered
5+
AI Models Deployed
15+
Clients Served

The best AI systems are not the most complex — they are the ones that solve real problems reliably, at scale, in production.

— My Engineering Philosophy

Currently Working On
  • 🤖 Advanced multi-modal RAG systems
  • 🌾 Expanding Zari.AI to new crop diseases
  • 🛡️ Publishing cyber defense research
Technical Arsenal

Core Skills & Technologies

A battle-tested toolkit for building production AI systems — from raw data to deployed intelligent applications.

Machine Learning & Deep Learning

PyTorchTensorFlowScikit-LearnComputer VisionOpenCVResNeXtYOLOCNNsLSTMReinforcement Learning

Generative AI & NLP

TransformersLLM Fine-TuningLoRA / QLoRARAG PipelinesLangChainLlamaIndexVector DatabasesPrompt EngineeringHuggingFaceOpenAI API

MLOps, LLMOps & Infrastructure

DockerFastAPIFlaskGitLinuxModel DeploymentModel MonitoringCI/CDREST APIsWhatsApp APILLM EvaluationGuardrailsPrompt VersioningLLM ObservabilityLangSmith

Languages & Tools

PythonC++SQLGoogle ColabJupyterPandasNumPyMatplotlibStreamlitPostman

Core Proficiency

Python & ML Frameworks92%
Generative AI & LLMs88%
MLOps & LLMOps82%
Featured Work

Production AI Projects

Real-world intelligent systems built from research to deployment — spanning computer vision, generative AI, and autonomous agents.

Featured Project

Zari.AI

Agricultural Disease Detection

Voice-enabled AI system providing real-time crop disease diagnosis via web and WhatsApp interfaces. Helps farmers in rural areas detect diseases and get actionable treatment recommendations instantly through natural conversation.

PythonComputer VisionSpeech ProcessingFastAPIWhatsApp APIPyTorch

Key Features

  • Real-time disease detection via camera input
  • Multi-language voice interface
  • WhatsApp integration for field accessibility
  • FastAPI backend with sub-200ms response
Research Project

Deepfake Detection Platform

ResNeXt + LSTM Pipeline

End-to-end deepfake video verification pipeline combining ResNeXt spatial feature extraction with LSTM temporal analysis. Includes a Flask backend API and a Chrome Extension for real-time browser-based verification of media authenticity.

PyTorchResNeXtLSTMFlaskChrome Extension APIOpenCV

Key Features

  • Dual-stream ResNeXt + LSTM architecture
  • Chrome Extension for instant web verification
  • 90%+ detection accuracy on benchmark datasets
  • Frame-level confidence scoring with heatmaps
Systems AI

Autonomous Cyber Defense System

Multi-Agent RL for Network Security

Hierarchical multi-agent reinforcement learning pipeline for real-time network threat detection and automated mitigation. Agents autonomously identify intrusion patterns, classify attack types, and deploy countermeasures without human intervention.

PythonPyTorchMulti-Agent RLNetwork AnalyticsGymnasium

Key Features

  • Hierarchical MARL with strategic + tactical agents
  • Real-time anomaly detection & classification
  • Autonomous threat mitigation protocols
  • Network graph-based state representation
Career & Milestones

Experience & Journey

My professional path as an AI Engineer, Studio Founder, and Machine Learning researcher.

icode Studios — <you imagine, WE code> official logo
Workicode Studios — <you imagine, WE code>

Founder & CEO

icode Studios — <you imagine, WE code>

Pakistan & Remote2025 — Present

Founded and leading icode Studios — an AI engineering & custom software development studio delivering production AI systems, LLM pipelines, and full-stack applications.

  • Leading AI product strategy, system architecture & client solution design
  • Building custom RAG architectures, LLM fine-tuning & agentic workflows
  • Delivering scalable web, mobile & cloud software for global clients
University of Haripur official logo
EducationUniversity of Haripur

BS Artificial Intelligence

University of Haripur

Haripur, Pakistan2022 — 2026

Completed Bachelor of Science in Artificial Intelligence focusing on Computer Vision, Deep Learning, and real-world AI applications.

  • Completed Deepfake Detection System combining ResNet + LSTM architectures
  • Object Detection projects using YOLO (Number Plate & Smartphone Detection)
  • Specialized in Deep Learning, Computer Vision & Pattern Recognition
Work

Machine Learning Engineer Intern

Avant Labs

Islamabad, Pakistan2024 — Present

Developed and trained Machine Learning models and Convolutional Neural Networks (CNNs), while gaining hands-on introduction to mobile integration using Flutter.

  • Trained & optimized Machine Learning models & CNNs
  • Gained hands-on experience with Flutter mobile app integration
  • Worked on model inference and backend pipeline optimization
Work

Data Entry & Data Visualization Intern

Galant Engineers

Islamabad, Pakistan2023 — 2024

Managed complex data entry, data cleaning, statistical analysis, and interactive visual dashboards for engineering workflows.

  • Cleaned, structured, and validated large-scale datasets
  • Created interactive data visualizations and executive dashboards
  • Streamlined reporting workflows for engineering analytics
Work

Freelance AI Engineer

Upwork & Fiverr

Remote — Worldwide2023 — Present

Delivered end-to-end AI and computer vision solutions for international clients on Upwork and Fiverr.

  • Built custom object detection & NLP pipelines for clients
  • Developed AI-powered chatbots & intelligent backends
  • Delivered production-ready ML solutions globally
GIKI — Ghulam Ishaq Khan Institute official logo
EducationGIKI — Ghulam Ishaq Khan Institute

Advanced AI Bootcamp 2026

GIKI — Ghulam Ishaq Khan Institute

Topi, Swabi, Pakistan2026

Attended an intensive Advanced AI Bootcamp at GIKI covering cutting-edge Generative AI, RAG, Large Language Models, and production deployment pipelines.

  • Full RAG (Retrieval-Augmented Generation) & NLP Architectures
  • Deep Learning & LLM Fine-Tuning (LoRA / QLoRA)
  • End-to-End MLOps & LLMOps Pipeline Engineering
Verified Credentials

Certifications & Specializations

Continuous learning credentials from leading AI organizations including Anthropic, IBM, DeepLearning.AI, and Vanderbilt University.

Jul 2026

Certificate of Completion: Claude 101

Anthropic

ID: rqyino9f2azv
Claude AIPrompt EngineeringLLM APIs
Jul 2026

AI Fluency Framework & Foundations

Anthropic

ID: 9p7xdnwknfz6
Prompt EngineeringAI Foundations
Jun 2025

Develop Generative AI Applications: Get Started

IBM

ID: JYEPO77WHTAV
Generative AIApplication DevelopmentPython
Jun 2025

Build RAG Applications: Get Started

IBM

ID: 7RFD0U06D9A1
RAG PipelinesVector DatabasesLangChain
Feb 2025

Generative AI and Model Selection

Vanderbilt University

ID: V5YQQLJLG755
Model SelectionLLM EvaluationFine-Tuning Strategy
Feb 2025

AI For Everyone

DeepLearning.AI

ID: P9ZL6GAT5SSJ
AI StrategyMachine Learning ConceptsAI Ethics
Sep 2023

Hack N' Connect

Ghulam Ishaq Khan Institute (GIKI)

Computer NetworkingSystems
2026

AI Agents in LangGraph (Short Course)

Coursera / DeepLearning.AI

LangGraphAgentic WorkflowsMulti-Agent Systems
2026

Python for Data Science, AI & Development

IBM / Coursera

PythonPandasData ScienceAPI Integration
2026

Foundations of Cybersecurity

Google / Coursera

CybersecurityNetwork SecurityThreat Analysis
Get In Touch

Let's Build Something Together

Whether you have a project in mind, want to collaborate, or just want to say hi — my inbox is always open.

Location

Pakistan 🇵🇰 — Available Remotely

Response Time

Usually within 24 hours