Abdul
Rahman

I build intelligent systems that work in the real world. As an AI Researcher & Engineer, I design scalable ML pipelines, LLM-powered applications, and visual analytics systems, bridging ideas with robust, product-grade execution.

View Work

Get in Touch

Tech Stack

AI & Machine Learning

Python PyTorch HuggingFace LangChain Scikit-learn Keras OpenCV

Data & Infrastructure

SQL Pandas NumPy AWS Docker MLflow PostgreSQL Kubernetes

Focus: RAG pipelines, agentic workflows, model optimization, fine-tuning (LoRA/QLoRA), and production MLOps.

Multi-Agent System

snippet

from langchain.agents import Tool, AgentExecutor
from langchain.chat_models import ChatOpenAI
from autogen import AssistantAgent, UserProxyAgent

# Define specialized agents
researcher = AssistantAgent(
  name="researcher",
  llm_config={"model": "gpt-4", "temperature": 0},
  system_message="You search and extract findings."
)

analyst = AssistantAgent(
  name="analyst",
  llm_config={"model": "gpt-4", "temperature": 0},
  system_message="You synthesize insights."
)

# Orchestrate multi-agent workflow
coordinator = UserProxyAgent(name="coord")

def run(query: str) -> str:
  coordinator.initiate_chat(researcher, msg=query)
  findings = researcher.last_message()
  return analyst.reply(findings)

Timeline

2018 — Present

Researcher — Northern Illinois University

Building production ML systems across healthcare, environmental monitoring, and social media. Publishing at IEEE VIS and TVCG.
2020 — Present

Lab Head — NIU VA Lab

Leading AI-driven visualization research, mentoring 5 graduate students, advancing multi-view exploratory analytics.
2016 — 2017

Data Analyst — Amazon

Optimized e-commerce databases by 30%, built data quality workflows validating 500K+ daily records.

Impact

Production systems and published research with measurable outcomes.

Performance Gains

Inference

4×

Throughput

GPU Memory

-60%

LoRA tuning

Training

-45%

Time reduced

Scale & Systems

Records validated daily 500K+

Images processed 300K+

Data accuracy 91%+

Query latency reduced 30%

Research Output

Publications 9 papers

Top venues IEEE VIS, TVCG

Recognition Best Paper

Program Committee CIKM, JCDL

Domains & Expertise

Healthcare AI FinTech E-commerce RAG Systems Multi-Agent LLMs LLM Fine-tuning Multimodal AI Computer Vision Data Visualization HCI

About

AI/ML Engineer finishing a Ph.D. at NIU. I've shipped ML systems at Amazon, won Best Paper at IEEE VIS, and now build LLM infrastructure that scales, from RAG pipelines to multi-agent systems.

Years

Papers

Students Mentored

What I am building

Building LLM systems end-to-end - fine-tuning, architecture optimization, and real-time inference at scale. Currently exploring transformer internals and memory architectures, while my dissertation explores how LLMs can reshape multi-view data exploration.

Get in touch

Research

Publications, preprints, and ongoing research in AI, machine learning, and data visualization.

Publications

90+

Citations

In Review

Featured Published Scientometrics 2023

YouTube and Science: Models for Research Impact

Abdul Rahman Shaikh, Hamed Alhoori, M. Sun

Investigates the growing number of videos citing scholarly articles, introduces new datasets, and builds ML models to predict citations, popularity, and public engagement using altmetrics signals.

Paper Code

Published Graphics Interface 2025

iTrace: Interactive Tracing of Cross-View Data Relationships

Abdul Rahman Shaikh, Maoyuan Sun, Xingchen Liu, Hamed Alhoori, David Koop

Presents iTrace, an interaction technique that guides attention with smooth focus transitions to make cross-view relationship tracing easier when related items are far apart and visually complex.

Paper Code

Published IEEE VIS 2022

Toward systematic design considerations of organizing multiple views

Abdul Rahman Shaikh, David Koop, Hamed Alhoori, Maoyuan Sun

Reviews multiple-view visualization systems and identifies layout considerations grounded in perception and content, explaining how spatial organization affects connecting information across views.

Paper

Published TVCG 2021 🏆 Best Paper

SightBi: Exploring Cross-View Data Relationships with Biclusters

Maoyuan Sun, Abdul Rahman Shaikh, Hamed Alhoori, Jian Zhao

Introduces SightBi, a visual analytics approach that formalizes cross-view relationships as biclusters and creates stand-alone relationship-views to guide exploration and reduce trial-and-error.

Paper Code

Published JCDL 2025

Generation, Evaluation, and Explanation of Novelists' Styles with Single-Token Prompts

Mosab Rezaei, Mina Rajaei Moghadam, Abdul Rahman Shaikh, Hamed Alhoori, Reva Freedman

A stylometry framework that fine-tunes LLMs with minimal single-token prompts to generate 19th-century novelist styles, and evaluates them using a transformer-based detector plus explainable analyses.

Paper Code