About
Senior Machine Learning Engineer & Data Scientist.
- Phone: +14026083435
- City: Raleigh, NC
- Degree: PhD
- Email: shamsbasir@gmail.com
AWS Certified Senior Machine Learning Engineer specializing in Generative AI with 6+ years of experience in machine learning, deep learning, and AI product development. Expert in LLMs, Generative AI (GANs, VAEs, Diffusion Models), NLP , and building scalable AI-driven solutions. Proficient in Python, MLOps, LLMOps, CI/CD, Git , and data science frameworks such as Scikit-Learn, NumPy, and Pandas. Skilled in training, fine-tuning, and deploying machine learning models on AWS cloud platforms, and creating user-centric dashboards using Streamlit . Demonstrates a proven track record of developing and deploying innovative chatbots and conversational agents , including LLM-powered solutions for clinical and commercial applications. Delivers scalable, user-aware AI systems that streamline workflows, enhance customer engagement, and generate significant business value.
Skills
Resume
Summary
Shamsulhaq Basir
- Raleigh, NC
- shamsbasir@gmail.com
Education
PhD in Mechanical Engineering (Machine Learning)
Aug 2018 - April 2023
University of Pittsburgh, Pittsburgh, PA
Specialization: Physics-informed Machine Learning (ML)
Coursework: GPA 4.0/4.0
- Parallel Computing (MPI, CUDA, OpenACC)
- Numerical Methods in Scientific Computing
- Linear and Complex Analysis
- Optimization Methods
- Reduced Order Modeling
- Advanced Fluid Mechanics
Visiting Student - Machine Learning
Jan 2020 - Jan 2022
Carnegie Mellon University, Pittsburgh, PA
Coursework: GPA 4.0/4.0
- Machine Learning 10601
- Deep Learning 11-785
- Machine Learning with Large Data Sets 10-605
Bachelor of Science - Civil Engineering (Double Major)
2014 - 2017
Middle East Technical University, Ankara, Turkey
Coursework: GPA 3.30/4.0
- Calculus of Functions of Several Variables
- Object Oriented Programming with C/Fortran
- Structural Analysis, Soil Mechanics
- Introduction to Probability and Stat I, tatistical Methods For Engineers
- Principles of Transportation Traffic
Bachelor of Science - Aerospace Engineering
2012 - 2017
Middle East Technical University, Ankara, Turkey
Coursework: GPA 3.50/4.0
- Calculus and Analytic Geometry, Differential Equations
- Thermodynamics, Aerodynamics, Propulsion
- Heat and Mass Transfer, Flight Mechanics, Mechanical Vibration
- Aerosace Engineering Design
Professional Experience
Senior Machine Learning Engineer
April 2024 - Present
Quinstreet, Remote
- As the sole Full-Stack Machine Learning Engineer and AI Product Specialist, lead the end-to-end development lifecycle of Large Language Model (LLM)-powered AI Agents and Chatbots
- Architect re-engagement solutions using voice, SMS, and email assistants to enhance customer interactions and optimize monetization
- Designed, developed and deployed two LLM-powered chatbots boosting engagement by 20% and generating $100,000 USD monthly revenue
- Architected GPT-4o-mini and LLM-based chatbots with advanced intent classification and automatic language detection
- Integrated RAG framework with advanced query optimization, delivering high-accuracy responses
Senior Manager, Data Science
Jan 2023 - April 2024
Bristol Myers Squibb, Remote
- Led NLP and Generative AI initiatives, specializing in question-answering chatbot development using PyTorch and Hugging Face
- Clinical Trial Design Intelligence Chatbot (ChatCTDI): Generated $5 million in annual savings through advanced question-answering platform
- Line of Therapy (LoT) Adjudication Solution: Achieved 93% accuracy in cancer therapy lines adjudication (pending patent)
- Developed AI-Powered Data Review Assistant for data inconsistency detection improving efficiency by 20x
Data Scientist
June 2022 - Aug 2022
Bristol Myers Squibb, Remote
- Clinical Trial Analysis: Collaborated on Upper Gastrointestinal and prostate cancer trials using ADaM and SDTM data formats
- Developed data factory using AWS Redshift, Python, and Pandas for cancer data analysis
- Investigated heterogeneous treatment effects using deep learning on time-to-event data
Graduate Student Researcher (Machine Learning)
Aug 2018 - June 2022
University of Pittsburgh, Pittsburgh, PA
- Physics-Informed Machine Learning: Led groundbreaking research in physics-informed neural networks (PINNs) and domain decomposition methods, publishing in top-tier journals
- Failure Modes in PINNs: Investigated failure modes of PINNs, uncovered scale-related challenges and proposed solutions
- Physics and Equality Constrained Neural Networks (PECANNs): Engineered a framework aligning neural networks with physics laws and multi-fidelity data
- Domain Decomposition with Neural Networks: Explored domain decomposition methods using AI to solve partitioned domains efficiently, developing parallel numerical methods with MPI
Certifications
Projects
- All
- Scientific ML
- Deep Learning
Publications
Journal Publications
Non-overlapping, Schwarz-type domain decomposition method for physics and equality constrained artificial neural networks
Computer Methods in Applied Mechanics and Engineering (2024): 117706
Authors: Hu, Qifeng, Basir, Shamsulhaq, and Senocak, Inanc
Investigating and Mitigating Failure Modes in Physics-Informed Neural Networks (PINNs)
Communications in Computational Physics 33.5 (2023): 1240–1269
Authors: Basir, Shamsulhaq
Physics and Equality Constrained Artificial Neural Networks: Application to Forward and Inverse Problems with Multi-fidelity Data Fusion
Journal of Computational Physics (2022): 111301
Authors: Basir, Shamsulhaq, and Senocak, Inanc
Preprints
A Generalized Schwarz-type Non-overlapping Domain Decomposition Method using Physics-constrained Neural Networks
arXiv preprint (2023)
Authors: Basir, Shamsulhaq, and Senocak, Inanc
arXiv: 2307.12435
An Adaptive Augmented Lagrangian Method for Training Physics and Equality Constrained Artificial Neural Networks
arXiv preprint (2023)
Authors: Basir, Shamsulhaq, and Senocak, Inanc
arXiv: 2306.04904
Conference Papers
Critical Investigation of Failure Modes in Physics-informed Neural Networks
AIAA SCITECH 2022 Forum
Authors: Basir, Shamsulhaq, and Senocak, Inanc
DOI: 10.2514/6.2022-2353
3
Journal Papers
2
Preprints
1
Conference Paper
144+
Citation
Talks & Presentations
| Date | Title | Event | Links |
|---|---|---|---|
| 2023 | Navigating the Landscape of Large Language Models (LLMs) for Clinical Research | BioTechX - US | Slides |
| 2023 | Fixing Physics-informed Neural Networks (PINNs) | Brown University | Slides |
| 2022 | Physics Mining: Teaching Machines To Obey Physics | University of Pittsburgh | Slides |
| 2022 | Physics and Equality Constrained Artificial Neural Networks | Brown University | Slides |
| 2022 | Physics and Equality Constrained Artificial Neural Networks | Alan Turing Institute | Slides |
| 2022 | Physics-informed and Equality-constrained Artificial Neural Networks | American Physics Society (APS) | Slides |
| 2022 | Efficient and Robust Training Strategies for Physics and Equality Constrained Artificial Neural Networks | American Physics Society (APS) | Slides |
| 2022 | Critical Investigation of Failure Modes in Physics-informed Neural Networks | AIAA SCITECH | Slides |
| 2021 | Parallel solution of Partial Differential Equations on Binarized Octrees | American Physics Society (APS) | Slides |
Martial Arts Journey
BJJ Double Gold Medalist 🥇🥇 | More videos coming soon from Muay Thai, MMA, and Boxing training!
Subscribe to my YouTube channel for more content: @ShamsBasir
First Competition Match
No-Gi - 2024
Won by Guillotine Choke
Second Competition Match
No-Gi - 2024
Won by Guillotine Choke
Third Competition Match
No-Gi - 2024
Won by Guillotine Choke
First Blue Belt Competition Match
No-Gi - 2024
Won by Guillotine Choke
First Round - IKF Advanced Muay Thai Match
IKF - 2024
Hurt him early (He was 10 pounds heavier too. Check my channel for the full story).
Second Round - IKF Advanced Muay Thai Match
IKF - 2024
Hurt and dropped my opponent (for the full story check out my channel.)