Kaleem Ullah Qasim
Experience
AI Engineer & LLM Specialist
Upwork
- • Top Rated freelancer on Upwork (top 10%), with a 100% job success rate and all 5-star ratings from 20+ clients.
- • Provided AI solutions using LangChain, LlamaIndex, CrewaI, and prompt engineering, increasing efficiency by 20%.
- • Optimized local LLMs for privacy-focused tasks, ensuring data security while improving task accuracy by 25%.
- • Developed modular AI tools with LangChain and Streamlit, boosting data retrieval accuracy by 35%.
- • Integrated multimodal AI systems for analytics, enhancing decision-making and achieving 100% client satisfaction.
Researcher - Traffic Management via LLMs Reasoning
Dr. Tariq Alsahfi (University of Jeddah)
- • Published research papers on LLM-based traffic flow and accident severity prediction in high-impact journals.
- • Developed the TraffiCoT framwork, improving traffic prediction accuracy by 19% over ML and 23% over DL models.
- • Reduced paper acceptance timelines by 30% through streamlined research and writing workflows.
- • Collaborated on projects integrating LLMs with geographic data to enhance traffic analysis and optimization.
Researcher - AI Security & Cyber Deception
Dr. Haitao Xu (Zhejiang University)
- • Published high-impact papers on AI-driven cybersecurity, focusing on deceptive ads and affiliate marketing detection.
- • Developed the AdsFlow algorithm for ad detection and classification, improving ad analysis accuracy.
- • Built Chrome extensions and tools for real-time deceptive ad detection and URL spoofing prevention.
- • Designed an ads intention classification system to identify deceptive intentions, enhancing fraud detection.
- • Applied NLP techniques to streamline e-crime investigations, reducing analysis time by 40%.
Data Scientist
Chengdu Ayurveda Biotechnology Co., Ltd China
- • Led the company to #1 position on Alibaba's medical equipment marketplace in China, achieving 180% YoY growth in sales.
- • Optimized SEO and search visibility, resulting in 95% increase in product search appearances.
- • Increased company revenue by 35% through predictive analytics and trend forecasting of medical equipment sales.
- • Developed ML models for inventory optimization, reducing stockout incidents by 45% and storage costs by 28%.
- • Implemented real-time sales analytics dashboard, improving decision-making response time by 60%.
- • Created customer segmentation model that increased targeted marketing efficiency by 40%.
Publications
Recursive Decomposition of Logical Thoughts:Framework for Superior Reasoning and Knowledge Propagation in LLMs
Kaleem Ullah Qasim, Jiashu Zhang, Tariq Alsahfi, Ateeq Ur Rehman Butt
Journal of Artificial Intelligence Research (JAIR), 2024
From data to decisions: enhancing financial forecasts with LSTM for AI token prices
Kaleem Ullah Qasim, Rizwan Ali, Muhammad Waqas Aslam
Journal of Economic Studies, 2023
Understanding the Business of Online Affiliate Marketing: An Empirical Study
Haitao Xu, Yiwen Sun, Kaleem Ullah Qasim
IEEE INFOCOM 2025 Main Conference, 2024
Projects
NL2SQL Intention Aware Agent
- • Developed an NL2SQL chatbot for querying tabular data
- • Added intent classification for routing queries
- • Enabled SQL generation, execution, and visualizations
- • Follow-up question generation for user engagement
- • Simplified data interaction for non-SQL users
SaRGeN (Suspicious Activity Report Generator)
- • Suspesious transaction detection and reporting
- • Automated SAR filing for financial institutions
- • Real-time transaction monitoring and alerts
- • Dynamic visualization of transaction patterns
- • Integrated Groq for real-time high speed inference
Real-time Civility Enforcement Engine
- • Real-time toxicity detection in chat messages
- • Automated moderation and user warnings
- • Dynamic visualization of toxicity trends
- • Integrated Detoxify for real-time inference
- • Reduced toxic content by 100% in user interactions
PaperPat - arXiv Paper Search Engine
- • Secure authentication with encrypted storage.
- • Advanced search with filters.
- • Bulk downloads with metadata handling.
- • Responsive UI with 60% faster load times.
- • Reliable SQLite integration with 99.9% uptime.
Education
Ph.D. in Artificial Intelligence
Southwest Jiaotong University - China
2022 - Present
Specialization: Reasoning in LLMs, Prompt Engineering
Master in Computer Application Technology
Southwestern University of Finance and Economics - China
2019 - 2022
Specialization: NLP, Machine Learning, NLU, NLI
Mandarin Certification
Beijing Language and Culture University - China
2016 - 2018
Specialization: Chinese Language Proficiency, Chinese Culture
Bachelor of Information Technology
University of Education - Pakistan
2011 - 2015
Specialization: Software Engineering