Building intelligent systems with ML, automation & reproducible pipelines.
Hands-on experience in Python-based data pipelines, MLOps, LLM fine-tuning, and AI automation workflows. Specializing in reproducible ML systems, from data ingestion to deployment-ready artifacts.
Reproducible ML
DVC + Git for data/model versioning, pinned environments, deterministic runs.
Pipeline Automation
Airflow DAGs and CI/CD to remove manual steps and prevent regressions.
Deployment-ready
Dockerized training/inference and cloud-ready artifacts with S3-backed storage.
About
Practical ML engineering: real projects, reproducible pipelines, deployable artifacts.
About
Aspiring AI Engineer
I'm an aspiring AI Engineer and a senior B.Sc. student in Artificial Intelligence at FAST-NUCES (Aug 2022 – Jun 2026), with hands-on experience building and shipping ML workflows.
From Python-based data pipelines and preprocessing to AI automation workflows and LLM fine-tuning, I focus on engineer-grade deliverables: clear repository structure, deterministic runs, environment pinning, and deployment-ready artifacts.
At Nexium (Jun–Aug 2025), I worked as an AI Workflow Automation Intern, integrating LLMs using LangChain with MongoDB and Supabase, automating workflows via n8n, and managing code in Git — enabling web apps to handle requests automatically and reducing manual effort.
What I optimize for
- Reproducibility over one-off runs
- Automation over manual runbooks
- Artifacts over screenshots
- Clear inputs/outputs per step
Core toolkit
- GitHub (reviews, CI, releases)
- Docker (portable runtime)
- MLflow & DVC (versioning)
- Airflow (orchestration)
- AWS / Azure (cloud)
Machine Learning & AI
Hands-on experience with AI Agent Workflows, Computer Vision, NLP, Recommender Systems, and Deep Q-Learning for real-world applications.
AI Automation & Workflows
Skilled in building AI-powered automation using n8n, LangChain, and LLM integrations with MongoDB, Supabase, and cloud services.
Data Pipelines & Engineering
Proficient in Python-based data pipelines, EDA, feature engineering, SQL, and data version control for scalable ML solutions.
MLOps & Deployment
Experienced with Docker, Git, GitHub Actions, MLflow, Apache Airflow, AWS SageMaker, and Microsoft Azure for model deployment and monitoring.
Core Skills
End-to-end AI development — from research and model training to production deployment.
AI Agent Workflows & LLMs
Hands-on experience with AI Agent Workflows, Hugging Face Transformers, pretrained LLMs, model fine-tuning, and RAG with LangChain for building intelligent applications.
Computer Vision & NLP
Skilled in Computer Vision, Natural Language Processing, Recommender Systems, and Deep Q-Learning. Built projects spanning image classification, text analysis, and reinforcement learning.
Data Science & Engineering
Proficient in SQL, EDA, Feature Engineering, and building end-to-end data pipelines. Experienced with data gathering, cleaning, and preprocessing for ML workflows.
MLOps & Deployment
Experienced with Docker, Git, GitHub Actions, MLflow & Neptune, Data Version Control, and Apache Airflow for reproducible ML workflows and CI/CD pipelines.
Cloud Platforms
Hands-on with AWS SageMaker and Microsoft Azure for model training, deployment, and scaling. Familiar with cloud-native ML services and infrastructure.
AI Automation & Integration
Skilled in n8n workflow automation, LangChain integrations with MongoDB and Supabase, and building AI-powered web applications that handle requests automatically.
Technology Stack
Tools and frameworks I use to build, train, and deploy ML systems.
Machine Learning & AI
Data Science & Engineering
MLOps & DevOps
Cloud & Infrastructure
Interested in Working Together?
Let's discuss how my skills in ML, AI automation, and data engineering can contribute to your team.
Certifications
Professional credentials from leading institutions in AI and machine learning.
Machine Learning Specialization
DeepLearning.AI
Comprehensive specialization covering machine learning fundamentals, neural networks, and deep learning techniques for real-world applications.
Verify CertificateGen AI Foundational Models for NLP & Language Understanding
IBM
Advanced course on foundational models for natural language processing, covering transformer architectures and language understanding.
Verify CertificateGenerative AI Engineering and Fine-Tuning Transformers
IBM
Hands-on course on engineering generative AI systems and fine-tuning transformer models for specific use cases.
Verify CertificateGenerative AI and LLMs: Architecture and Data Preparation
IBM
Comprehensive course covering LLM architectures, data preparation techniques, and implementation strategies for generative AI systems.
Verify CertificateResearch & Development
Academic research and ML experiments demonstrating deep technical expertise.
AI-Powered Urdu News Transcription & Archiving
Final Year Project at FAST-NU — Built an advanced AI-powered transcription system for Urdu news archives, converting video content into precise, searchable, and editable text with high-accuracy timestamped outputs.
Result: High-accuracy timestamped outputs with scalable storage
Impact: Preserves Pakistan’s broadcast heritage through intelligent archiving and tagging technology
Deep Reinforcement Learning for Traffic Signal Control
Implemented DQN-based traffic signal control trained in SUMO simulator, using Deep Reinforcement Learning for adaptive signal timing to optimize urban traffic flow.
Result: ~80% reduction in vehicle waiting time vs fixed-time controllers
Impact: Demonstrated AI-driven urban traffic optimization with real-world simulation
RAG-Based Customer Support Agent
Built PromptDesk — an AI-powered customer support agent using LangChain and OpenAI GPT with semantic search over internal FAQs, documents, and past tickets.
Result: Semantic search with source citations and confidence scoring
Impact: End-to-end RAG pipeline with human escalation for low-confidence responses
Multimodal AI Chatbot System
Developed MultiSense Agent — a multimodal chatbot capable of processing text, voice, images, and PDFs using RAG and n8n automation, with WhatsApp integration.
Result: Supports text, voice, image, and PDF processing in real-time
Impact: Scalable multi-format AI assistant deployed with Docker containers
Supervised ML for Disease Prediction
CardioML — Leveraged various supervised classification algorithms (SVC, Naive Bayes, Decision Trees, Random Forest, Logistic Regression) to predict disease likelihood.
Result: Comparative analysis across 5+ classification algorithms
Impact: Applied ML to healthcare domain with systematic model comparison and evaluation
Cryptocurrency Data Analysis Pipeline
TokenScope — Built a data analysis pipeline that fetches live cryptocurrency data via APIs, cleans and processes it using Pandas, and generates insights through statistical analysis.
Result: Real-time trend reports and volatility analysis
Impact: End-to-end data pipeline from API ingestion to visualization and insight delivery
Projects
Technical projects across AI, automation, and data engineering.
Pak Journal Archive System
FAST-NU Final Year Project — Preserves Pakistan’s broadcast heritage through precise AI transcription and intelligent archiving. Advanced AI-powered transcription system for Urdu news archives, converting video content into precise, searchable, and editable text with high-accuracy timestamped outputs and scalable storage.
MultiSense Agent
Multimodal AI chatbot with WhatsApp integration capable of processing text, voice, images, and PDFs using RAG and n8n automation. Seamlessly interacts with users via WhatsApp API, supports multiple data formats, and is deployed with Docker for scalable, reliable performance.
PromptDesk
AI-powered customer support agent built with Python, LangChain, OpenAI GPT, a vector database, FastAPI, and React. Answers user queries using semantic search over internal FAQs, documents, and past tickets, with source citations and human escalation for low-confidence responses.
IntelliSignal-RL
DQN-based traffic signal control trained in SUMO, using Deep Reinforcement Learning for adaptive signal timing. Cuts vehicle waiting time by ~80% vs fixed-time controllers, showcasing AI-driven urban traffic optimization.
CardioML
Leverages various supervised classification algorithms (SVC, Naive Bayes, Decision Trees, Random Forest, Logistic Regression, etc.) to predict the likelihood of different diseases based on multiple input features.
TokenScope
Data analysis pipeline that fetches live cryptocurrency data via APIs, cleans and processes it using Pandas, and generates insights through statistical analysis and visualizations. Delivered trend reports and volatility analysis using Matplotlib.
Contact
Open to opportunities, collaborations, and conversations about AI and technology.
Send a Message
Fill out the form and I'll get back to you as soon as possible.
What I'm Looking For
- AI/ML Engineering opportunities
- AI Automation & Workflow projects
- Data engineering collaborations
- Open source contributions
- Internships & entry-level roles