# Uzair Arif

> AI researcher and machine learning engineer based in Islamabad, Pakistan, available for remote roles worldwide.

Uzair Arif builds and researches multi-agent AI systems, agentic memory architectures, forecasting pipelines, retrieval systems, and production MLOps. His work combines machine learning engineering with research into cooperation, governance, memory reliability, and strategic risk in language-model agents.

## Core Links

- [Website](https://uzairlol.github.io/): Official portfolio and personal website
- [GitHub](https://github.com/uzairlol): Open-source projects and code repositories
- [LinkedIn](https://www.linkedin.com/in/uzairarif/): Professional background and research updates
- [Google Scholar](https://scholar.google.com/citations?user=sshNYq4AAAAJ&hl=en): Academic research publications and citations
- [Resume](https://uzairlol.github.io/assets/uzair_arif_resume.pdf): Full curriculum vitae in PDF format

## Focus Areas

- Multi-agent LLM systems and institutional governance
- Agentic memory and information-poisoning resilience
- Deep learning and time-series forecasting
- Hybrid retrieval and retrieval-augmented generation
- Production MLOps with FastAPI, Docker, MLflow, and DVC

## Featured Projects

- [ELICIT: Emergent LLM Institutions for Climate Treaties](https://github.com/uzairlol/ELICIT-fyp): Multi-agent simulation framework testing whether language-model agents can create and follow institutions for climate cooperation.
- [SEAM: Shared Evolving Agent Memory](https://github.com/uzairlol/seam): Benchmark for studying how shared memories, false lessons, and information poisoning spread through self-evolving agent populations.
- [Taxi Demand MLOps Pipeline](https://github.com/uzairlol/taxi-mlops): Production-oriented system for real-time NYC taxi-demand forecasting with FastAPI, Docker, drift monitoring, and CI/CD tests.
- [ECG5000 Anomaly Detection with TCN](https://github.com/uzairlol/ECG5000-TCN-Classification): PyTorch time-series pipeline for ECG anomaly detection using temporal convolutional networks and contrastive learning.
- [IBA RAG Chatbot](https://www.linkedin.com/posts/uzairarif_built-a-rag-chatbot-designed-to-centralize-activity-7419017492427325441-mq58): Hybrid retrieval-augmented generation assistant using FAISS, BM25, Reciprocal Rank Fusion, LangChain, and Groq.
- [LSTM Stock Prediction](https://github.com/uzairlol/LSTM_AAPL_Stock_Prediction): Deep learning recurrent neural network forecasting Apple (AAPL) stock prices.
- [Bitcoin Sentiment Prediction](https://github.com/uzairlol/Bitcoin-Sentiment-Prediction): Sentiment analysis and price movement modeling for cryptocurrency markets.

## Site Navigation

- [Home](https://uzairlol.github.io/): Main portfolio landing page
- [About Me](https://uzairlol.github.io/#about): Research journey, technical stack, and education
- [Projects](https://uzairlol.github.io/#projects): Full catalog of AI and ML engineering projects
- [Experience](https://uzairlol.github.io/#experience): Professional and research experience
- [Contact](https://uzairlol.github.io/#contact): Inquiries and contact channels
- [Sitemap](https://uzairlol.github.io/sitemap.xml): XML sitemap for search and indexing agents