CV

PhD researcher in LLM safety, agentic AI, and multilingual AI systems, with 8+ years of experience building production-scale ML and backend systems. Published at ICLR, PETS, AAAI, HyperText, COMPSAC, and ASONAM.

A PDF version will be linked here once uploaded to files/CV.pdf.

Education

Ph.D. in Computer Science, University of Texas at El Paso (UTEP) — Aug 2024 – present
- Combined CGPA 3.92 / 4.00 (including credits transferred from SIUC)
M.Sc. in Computer Science, University of Texas at El Paso (UTEP) — Aug 2024 – May 2025
- CGPA 3.80 / 4.00
Ph.D. coursework in Computer Science, Southern Illinois University Carbondale (SIUC) — Jan 2023 – Jul 2024
- CGPA 4.00 / 4.00 (24 credits, transferred to UTEP)
B.Sc. in Computer Science and Engineering, Bangladesh University of Engineering and Technology (BUET) — Mar 2009 – May 2016
- CGPA 3.35 / 4.00

Research Experience

Graduate Research Assistant — SUPREME Lab, UTEP and SIUC · Jan 2023 – present

Advised by Dr. Sajedul Talukder
Research areas: generative AI, LLM safety and alignment, agentic AI safety, social network security, user data privacy
Built a large-scale LLM safety evaluation pipeline — 1.2M+ samples across 100 languages — for multilingual jailbreak detection across the Llama, Gemma, Qwen, Mistral, GPT, Gemini, and Claude families
Designed an AEGIS-based harm scoring framework (S1–S23) for cross-lingual safety benchmarking
Developed benchmark frameworks and defenses for agentic LLM skill / MCP supply-chain security, including detection and mitigation of malicious tool integrations

Work in Progress

Cap-PermLLM: Capability-Aware Access Control for Agentic Large Language Models — CCS 2026
Detecting Privacy Drift as an Early-Warning Signal of Account Compromise and Data Exposure — CCS 2026
Scalable Benchmark for Multilingual (100 Languages) Safety Alignment for Large Language Models — NeurIPS 2026
Unmasking Citation Manipulation: A Feature-Driven and Transformer-Based Approach — ICWSM 2026
Who Said What, When, and Where? Agentic Contextual Interpretation of Social Media Posts with Large Language Models — ICWSM 2026

Research Impact

Publications: 20+ peer-reviewed papers across ICLR, PETS, HyperText, AAAI Symposium, COMPSAC, and ASONAM
Citations: 130 · h-index 7 · i10-index 3 (Google Scholar, April 2026)
Research areas: LLM safety, AI alignment, social network security
Reviewer and PC service: MSR 2027 (Junior PC), SMS Workshop (PC), ASONAM 2023–2025, ICWSM 2026 (PC), WWW 2026 (PC), PLOS ONE, SN Computer Science
Awards: National Project Champion, BSCCL 2013 ($1,000); Top-56 team, U.S. DOE CyberForce Competition 2023
Travel grants: $600 for paper registration, HyperText 2025

Teaching Experience

Graduate Teaching Assistant — UTEP and SIUC · Spring 2024, 2025, 2026

UTEP, Fall 2025 and Spring 2026 — Computer Networks: mentoring and grading graduate and undergraduate students
UTEP, Spring 2025 — Automata: mentoring and grading graduate and undergraduate students
SIUC, Spring 2024 — led machine learning / NLP labs and Linux programming; mentored and graded students

Work Experience

Senior Software Engineer — BRAC IT, Anwar Group (AES), REVE Systems · Dhaka, Bangladesh · Mar 2017 – Jan 2023 (REVE Systems 2017–2020; AES 2022–2023)

Built and deployed ML training, LLM fine-tuning, and large-scale evaluation pipelines bridging research and production systems
Led backend and NLP-focused development across ERP, chatbot, and enterprise analytics systems using scalable, distributed service architectures
Implemented end-to-end NLP pipelines for text classification, entity recognition, and language detection in Python, scikit-learn, and NLTK
Integrated and optimized multi-platform applications (Salesforce, Shopify, WordPress) over REST APIs
Containerized applications with Docker and deployed on Linux servers; CI/CD workflows and basic container orchestration

Software Engineer — Nascenia IT, Aegis Foundry · Dhaka, Bangladesh · Aug 2014 – Mar 2017

Developed and maintained REST APIs, data pipelines, and reporting modules for restaurant, healthcare, and e-commerce clients
Designed database schemas, user management, and payment gateway integrations for production systems
Built reporting tools for POS systems over large databases, with a focus on query optimization and backend performance

Technical Skills

LLM / generative AI: fine-tuning, RAG, agentic LLMs, prompt injection defense, multimodal GenAI, safety and alignment, content moderation, tool-augmented LLMs, LangChain, CrewAI
ML/AI stack: PyTorch, Transformers, scikit-learn, NLTK, pandas
Cloud and tools: AWS, Google Colab, Anaconda, Selenium, Spring MVC, Django, React.js
Databases and servers: MySQL, PostgreSQL, MongoDB, Nginx, Linux/CentOS
Languages: Python, Java, JavaScript

Publications

See the full publication list or my Google Scholar profile.

When Guardrails Fail: Multilingual Long-Context Safety in Tool-Using Agents

Md Jahangir Alam, Ismail Hossain, Tanzim Ahad, Sai Puppala, Sajedul Talukder (2026). "When Guardrails Fail: Multilingual Long-Context Safety in Tool-Using Agents." NeurIPS 2026, Long-Context Foundation Models (LCFM).

SafeCache: Security-Aware Context Eviction for Tool-Using Agents

Md Jahangir Alam, Tanzim Ahad, Ismail Hossain, S M Abrar Jahin, Sai Puppala, Sajedul Talukder (2026). "SafeCache: Security-Aware Context Eviction for Tool-Using Agents." Under review at MLSys.

PolyAEGIS: Benchmarking Large Language Model Safety Across 100 Languages

Md Jahangir Alam, Tanzim Ahad, Ismail Hossain, Sai Puppala, Syed Bahauddin Alam, Sajedul Talukder (2026). "PolyAEGIS: Benchmarking Large Language Model Safety Across 100 Languages." Under review at ICLR 2027.

Lost in Speech: Measuring the Multilingual Cross-Modal Safety Gap in Speech-Language Models

Md Jahangir Alam, Tanzim Ahad, Ismail Hossain, Sai Puppala, Sajedul Talukder (2026). "Lost in Speech: Measuring the Multilingual Cross-Modal Safety Gap in Speech-Language Models." Submitted to ACL Rolling Review (NAACL).

Guard-to-Action: Multilingual Long-Context Safety in Tool-Using Agents

Md Jahangir Alam, Ismail Hossain, Tanzim Ahad, Sai Puppala, Sajedul Talukder (2026). "Guard-to-Action: Multilingual Long-Context Safety in Tool-Using Agents." Under review at ICLR 2027.

Influence Is Not Authority: When Causal Guardrail Signals Make Legitimate Tool Use Look Like an Attack in Tool-Using LLM Agents

Tanzim Ahad, Ismail Hossain, Md Jahangir Alam, Sai Puppala, Syed Bahauddin Alam, Sajedul Talukder (2026). "Influence Is Not Authority: When Causal Guardrail Signals Make Legitimate Tool Use Look Like an Attack in Tool-Using LLM Agents." Preprint (arXiv).

When Safety Geometry Collapses: Fine-Tuning Vulnerabilities in Agentic Guard Models

Ismail Hossain, Sai Puppala, Jannatul Ferdaus, Md Jahangir Alam, Tanzim Ahad, Yoonpyo Lee, Syed Bahauddin Alam, Sajedul Talukder (2026). "When Safety Geometry Collapses: Fine-Tuning Vulnerabilities in Agentic Guard Models." AAAI 2026 Summer Symposium.

Semantic Intent Fragmentation: A Single-Shot Compositional Attack on Multi-Agent AI Pipelines

Tanzim Ahad, Ismail Hossain, Md Jahangir Alam, Sai Puppala, Yoonpyo Lee, Syed Bahauddin Alam, Sajedul Talukder (2026). "Semantic Intent Fragmentation: A Single-Shot Compositional Attack on Multi-Agent AI Pipelines." AAAI 2026 Summer Symposium.

HALLPERM: Exposing the Safety Illusion in LLM Tool Use via Implicit Privilege Escalation and Semantic Risk

Md Jahangir Alam, Tanzim Ahad, Ismail Hossain, Sai Puppala, Yoonpyo Lee, Syed Bahauddin Alam, Sajedul Talukder (2026). "HALLPERM: Exposing the Safety Illusion in LLM Tool Use via Implicit Privilege Escalation and Semantic Risk." AAAI 2026 Summer Symposium.

Agent-Fence: Mapping Security Vulnerabilities Across Deep Research Agents

Sai Puppala, Ismail Hossain, Md Jahangir Alam, Yoonpyo Lee, Jay Yoo, Tanzim Ahad, Syed Bahauddin Alam, Sajedul Talukder (2026). "Agent-Fence: Mapping Security Vulnerabilities Across Deep Research Agents." AAAI 2026 Summer Symposium.

SkillVetBench: LLM-as-Judge for Multi-Dimensional Security Risk Evaluation in Open-Source LLM Agent Skills

Ismail Hossain, Sai Puppala, Md Jahangir Alam, Tanzim Ahad, Sajedul Talukder (2026). "SkillVetBench: LLM-as-Judge for Multi-Dimensional Security Risk Evaluation in Open-Source LLM Agent Skills." Preprint (arXiv).

The Misattribution Gap: When Memory Poisoning Looks Like Model Failure in Agentic AI Systems

Tanzim Ahad, Ismail Hossain, Md Jahangir Alam, Sai Puppala, Syed Bahauddin Alam, Sajedul Talukder (2026). "The Misattribution Gap: When Memory Poisoning Looks Like Model Failure in Agentic AI Systems." Preprint (arXiv).

The Art of the Jailbreak: Formulating Jailbreak Attacks for LLM Security Beyond Binary Scoring

Ismail Hossain, Tanzim Ahad, Md Jahangir Alam, Sai Puppala, Syed Bahauddin Alam, Sajedul Talukder (2026). "The Art of the Jailbreak: Formulating Jailbreak Attacks for LLM Security Beyond Binary Scoring." Preprint (arXiv).

LAMDA: A Longitudinal Android Malware Benchmark for Concept Drift Analysis

Md Ahsanul Haque, Ismail Hossain, Md Mahmuduzzaman Kamol, Md Jahangir Alam, Suresh Kumar Amalapuram, Sajedul Talukder, Mohammad Saidur Rahman (2026). "LAMDA: A Longitudinal Android Malware Benchmark for Concept Drift Analysis." ICLR 2026.

AI-in-the-Loop: Privacy Preserving Real-Time Scam Detection and Conversational Scam-baiting by Leveraging LLMs and Federated Learning

Ismail Hossain, Sai Puppala, Md Jahangir Alam, Sajedul Talukder (2026). "AI-in-the-Loop: Privacy Preserving Real-Time Scam Detection and Conversational Scam-baiting by Leveraging LLMs and Federated Learning." PETS 2026.

Toward Empathetic AI: Neural-Symbolic LLMs for Emotionally Aligned Conversations

Ismail Hossain, Md. Jahangir Alam, Sai Puppala, Sajedul Talukder (2025). "Toward Empathetic AI: Neural-Symbolic LLMs for Emotionally Aligned Conversations." ASONAM 2025.

Real-Time Personalized Content Adaptation through Matrix Factorization and Context-Aware Federated Learning

Sai Puppala, Ismail Hossian, MD Jahangir Alam, Sajedul Talukder (2025). "Real-Time Personalized Content Adaptation through Matrix Factorization and Context-Aware Federated Learning." ASONAM 2025.

Privacy Control in Social Networks: Integrating Behavioral Patterns and Content Sensitivity for Audience Recommendation

Md Jahangir Alam, Ismail Hossain, Sai Puppala, Sajedul Talukder (2025). "Privacy Control in Social Networks: Integrating Behavioral Patterns and Content Sensitivity for Audience Recommendation." ASONAM 2025.

LLMs Against Digital Deviance: Scalable Hate Speech Detection in Low-Resource and Code-Mixed Social Media

Md Jahangir Alam, Ismail Hossain, Sai Puppala, Sajedul Talukder (2025). "LLMs Against Digital Deviance: Scalable Hate Speech Detection in Low-Resource and Code-Mixed Social Media." ASONAM 2025.

A Comprehensive Survey of Federated Learning for Edge AI: Recent Trends and Future Directions

Sai Puppala, Ismail Hossain, Jahangir Alam, Tanzim Ahad, Sajedul Talukder (2025). "A Comprehensive Survey of Federated Learning for Edge AI: Recent Trends and Future Directions." Preprint.

LLM-Guided Dynamic-UMAP for Personalized Federated Graph Learning

Sai Puppala, Ismail Hossain, Md Jahangir Alam, Tanzim Ahad, Sajedul Talukder (2025). "LLM-Guided Dynamic-UMAP for Personalized Federated Graph Learning." Preprint (arXiv).

Variational Gaussian Mixture Manifold Models for Client-Specific Federated Personalization

Sai Puppala, Ismail Hossain, Md Jahangir Alam, Sajedul Talukder (2025). "Variational Gaussian Mixture Manifold Models for Client-Specific Federated Personalization." Preprint (arXiv).

Beyond Transformers: Leveraging Large Language Models and Encoder-Decoder Architectures for Emotion Detection in Low-Resource Language

Md Jahangir Alam, Ismail Hossain, Sai Puppala, Sajedul Talukder (2025). "Beyond Transformers: Leveraging Large Language Models and Encoder-Decoder Architectures for Emotion Detection in Low-Resource Language." ASONAM 2025.

Comprehensive Privacy Risk Assessment in Social Networks Using User Attributes Social Graphs and Text Analysis

Md Jahangir Alam, Ismail Hossain, Sai Puppala, Sajedul Talukder (2025). "Comprehensive Privacy Risk Assessment in Social Networks Using User Attributes Social Graphs and Text Analysis." HyperText 2025.

EVOLVE-X: Embedding Fusion and Language Prompting for User Evolution Forecasting on Social Media

Ismail Hossain, Sai Puppala, Md Jahangir Alam, Sajedul Talukder (2025). "EVOLVE-X: Embedding Fusion and Language Prompting for User Evolution Forecasting on Social Media." Preprint (arXiv).

Securing SDN-Based Advanced Reactors Against DDoS Attacks Using Quantum-Inspired Ensemble Models

Jahangir Alam, Ismail Hossain, Sai Puppala, Kazuma Kobayashi, Samrendra Roy, Syed Alam, Sajedul Talukder (2025). "Securing SDN-Based Advanced Reactors Against DDoS Attacks Using Quantum-Inspired Ensemble Models." NPIC&HMIT 2025.

Secure Anomaly Detection in Advanced Reactor Through Quantum Federated Learning with Advanced Cryptography

Sai Puppala, Ismail Hossain, Jahangir Alam, Kazuma Kobayashi, Samrendra Roy, Syed Alam, Sajedul Talukder (2025). "Secure Anomaly Detection in Advanced Reactor Through Quantum Federated Learning with Advanced Cryptography." NPIC&HMIT 2025.

Proactive Cybersecurity Framework for Nuclear Power Plants Using Quantum Large Language Models

Ismail Hossain, Jahangir Alam, Sai Puppala, Kazuma Kobayashi, Samrendra Roy, Syed Alam, Sajedul Talukder (2025). "Proactive Cybersecurity Framework for Nuclear Power Plants Using Quantum Large Language Models." NPIC&HMIT 2025.

A Privacy-Aware Cyber Attack Detection Framework for Advanced Reactors Using Data Fusion and Quantum Deep Learning

Ismail Hossain, Sai Puppala, Jahangir Alam, Samrendra Roy, Kazuma Kobayashi, Syed Alam, Sajedul Talukder (2025). "A Privacy-Aware Cyber Attack Detection Framework for Advanced Reactors Using Data Fusion and Quantum Deep Learning." NPIC&HMIT 2025.

Optimus-Q: Utilizing Federated Learning in Adaptive Robots for Intelligent Advanced Reactor Operations Through Quantum Cryptography

Sai Puppala, Ismail Hossain, Jahangir Alam, Kazuma Kobayashi, Samrendra Roy, Syed Alam, Sajedul Talukder (2025). "Optimus-Q: Utilizing Federated Learning in Adaptive Robots for Intelligent Advanced Reactor Operations Through Quantum Cryptography." NPIC&HMIT 2025.

LLM-Powered Chatbot for Safety Compliance Assistance in Advance Reactors

Ismail Hossain, Sai Puppala, Jahangir Alam, Kazuma Kobayashi, Samrendra Roy, Syed Alam, Sajedul Talukder (2025). "LLM-Powered Chatbot for Safety Compliance Assistance in Advance Reactors." NPIC&HMIT 2025.

SAFARI: Self-regulAted Clustered FederAted Learning in a HeteRogeneous EnvIronment

Sai Puppala, Ismail Hossain, Md Jahangir Alam, Sajedul Talukder (2024). "SAFARI: Self-regulAted Clustered FederAted Learning in a HeteRogeneous EnvIronment." ICMLA 2024.

A visual approach to tracking emotional sentiment dynamics in social network commentaries

Ismail Hossain, Sai Puppala, Md. Jahangir Alam, Sajedul Talukder (2024). "A visual approach to tracking emotional sentiment dynamics in social network commentaries." ICWSM 2024 · Social Network Analysis and Mining.

SocialRec: User Activity Based Post Weighted Dynamic Personalized Post Recommendation System in Social Media

Ismail Hossain, Sai Puppala, Md. Jahangir Alam, Sajedul Talukder (2024). "SocialRec: User Activity Based Post Weighted Dynamic Personalized Post Recommendation System in Social Media." ASONAM 2024.

SocFedGPT: Federated GPT-Based Adaptive Content Filtering System Leveraging User Interactions in Social Networks

Sai Puppala, Ismail Hossain, Md Jahangir Alam, Sajedul Talukder (2024). "SocFedGPT: Federated GPT-Based Adaptive Content Filtering System Leveraging User Interactions in Social Networks." ASONAM 2024.

FLASH: Federated Learning-Based LLMs for Advanced Query Processing in Social Networks Through RAG

Sai Puppala, Ismail Hossain, Md. Jahangir Alam, Sajedul Talukder (2024). "FLASH: Federated Learning-Based LLMs for Advanced Query Processing in Social Networks Through RAG." ASONAM 2024.

EVOLVE: Predicting User Evolution and Network Dynamics in Social Media Using Fine-Tuned GPT-Like Model

Ismail Hossain, Sai Puppala, Md. Jahangir Alam, Sajedul Talukder (2024). "EVOLVE: Predicting User Evolution and Network Dynamics in Social Media Using Fine-Tuned GPT-Like Model." ASONAM 2024.

Combating Echo Chambers in Online Social Network by Increasing Content Diversity in Recommendation

Md Jahangir Alam, Ismail Hossain, Sai Puppala, Sajedul Talukder (2024). "Combating Echo Chambers in Online Social Network by Increasing Content Diversity in Recommendation." ASONAM 2024.

CAMERA: Context Based Emotion Detection Framework and Its Evaluation

Md. Jahangir Alam, Ismail Hossain, Sai Puppala, Sajedul Talukder (2024). "CAMERA: Context Based Emotion Detection Framework and Its Evaluation." ASONAM 2024.

Bangla Emotion Detection Dataset with an Extended Taxonomy and Its Evaluation

Md. Jahangir Alam, Md. Shohanur Rahman Shohan, Ismail Hossain, Sai Puppala, Sajedul Talukder (2024). "Bangla Emotion Detection Dataset with an Extended Taxonomy and Its Evaluation." ASONAM 2024.

4WHContext: A Context Based Hate Speech Detection Framework from Social Media Posts

Md. Jahangir Alam, Ismail Hossain, Sai Puppala, Sajedul Talukder (2024). "4WHContext: A Context Based Hate Speech Detection Framework from Social Media Posts." ASONAM 2024.

Generative AI like ChatGPT in Blockchain Federated Learning: use cases, opportunities and future

Sai Puppala, Ismail Hossain, Md Jahangir Alam, Sajedul Talukder, Jannatul Ferdaus, Mahedi Hasan, Sameera Pisupati, Shanmukh Mathukumilli (2024). "Generative AI like ChatGPT in Blockchain Federated Learning: use cases, opportunities and future." Preprint (arXiv).

SocialGuard: Bangla Text-Based Gender Identification for Enhancing Integrity in Social Networks

Md Jahangir Alam, Sultan Ahmed, Ismail Hossain, Sai Puppala, Zahidur Talukder, Sajedul Talukder (2024). "SocialGuard: Bangla Text-Based Gender Identification for Enhancing Integrity in Social Networks." COMPSAC 2024.

SCAN: A HealthCare Personalized ChatBot with Federated Learning Based GPT

Sai Puppala, Ismail Hossain, Md Jahangir Alam, Sajedul Talukder (2024). "SCAN: A HealthCare Personalized ChatBot with Federated Learning Based GPT." COMPSAC 2024.

SCALE: Self-Regulated Clustered FederAted LEarning in a Homogeneous Environment

Sai Puppala, Ismail Hossain, Md Jahangir Alam, Zahidur Talukder, Syed Bahauddin, Sajedul Talukder (2024). "SCALE: Self-Regulated Clustered FederAted LEarning in a Homogeneous Environment." COMPSAC 2024.

Advancements in Multimodal Social Media Post Summarization: Integrating GPT-4 for Enhanced Understanding

Md Jahangir Alam, Ismail Hossain, Sai Puppala, Sajedul Talukder (2024). "Advancements in Multimodal Social Media Post Summarization: Integrating GPT-4 for Enhanced Understanding." COMPSAC 2024.

ContextGPT: Predictive Monitoring of Nuclear Power Plant with Contextually-Informed Situational Understanding

Raisa Hossain, Farid Ahmed, Jahangir Alam, Sajedul Talukder, Syed Alam (2024). "ContextGPT: Predictive Monitoring of Nuclear Power Plant with Contextually-Informed Situational Understanding." ANS Transactions 2024.

Blockchain-Integrated Secure Framework for Enhanced E-Government Services

Sajedul Talukder, Md Jahangir Alam, Ismail Hossain, Sai Puppala (2023). "Blockchain-Integrated Secure Framework for Enhanced E-Government Services." Book chapter, Springer (Signals and Communication Technology).

Towards Addressing Identity Deception in Social Media using Bangla Text-Based Gender Identification

Sultan Ahmed, Md Jahangir Alam, Sajedul Talukder, Ismail Hossain (2023). "Towards Addressing Identity Deception in Social Media using Bangla Text-Based Gender Identification." ASONAM 2023.

Monitoring Dynamics of Emotional Sentiment in Social Network Commentaries

Ismail Hossain, Sai Puppala, Md Jahangir Alam, Sajedul Talukder (2023). "Monitoring Dynamics of Emotional Sentiment in Social Network Commentaries." ASONAM 2023.

Combating Identity Attacks in Online Social Networks: A Multi-Layered Framework Using Zero-Knowledge Proof and Permissioned Blockchain

Md Jahangir Alam, Ismail Hossain, Sai Puppala, Sajedul Talukder (2023). "Combating Identity Attacks in Online Social Networks: A Multi-Layered Framework Using Zero-Knowledge Proof and Permissioned Blockchain." ASONAM 2023.

Md Jahangir Alam

CV