About me

I am a Ph.D. student in Computer Science at the AI Institute, advised by Prof. Amit P. Sheth at the University of South Carolina. My research interests primarily lie in Agentic AI, Multimodal-AI, Generative-AI, Time Series Analysis, and Neurosymbolic-AI. I am particularly excited about the opportunities to apply my research interests to real-world problems such as rare-event prediction, anomaly detection, causal discovery, root cause analysis and event understanding in various domains. I am keen on exploring further domains and earning new knowledge to enhance my research skills, analytical thinking, and problem-solving. Additionally, I am deeply interested in collaborative, translational research that has the potential to make a broader and meaningful impact.

I am Actively Seeking for Full Time Research Scientist Opportunities

I am currently looking for exciting Full Time research opportunities in Artificial Intelligence, Agentic AI, Multimodal AI. If you are looking for a dedicated and passionate researcher, please feel free to reach out to me at jayakodc@email.sc.edu or chathurangijks@gmail.com.







Watch My Ph.D. Proposal Defense
▶

Announcements

  • 2026AssetOpsBench was accepted to KDD 2026 as a real-world benchmark for AI-driven task automation in industrial asset management.
  • 2026Co-leading the KDD 2026 tutorial, "Building Reliable Industrial Agents with MCP: A Hands-on AssetOpsBench Tutorial for AI-Driven Operations."
  • May 2026Started as an AI Research Intern at IBM Research, working on industrial AI agents, AssetOpsBench scenarios, embodied question answering, failure-metadata knowledge graphs, and automated evaluation pipelines.
  • 2026IndustryAssetEQA was accepted to ACL 2026, presenting a neurosymbolic operational intelligence system for embodied question answering in industrial asset maintenance.
  • 2026Paper accepted to CVPR 2026: "Where Should Knowledge Enter? A Layered Framework for Knowledge Infusion in Multimodal Iterative Generative Models."
  • 2026CausalPulse and CausalTrace were accepted to AAAI-MAKE 2026 and AAAI-IAAI 2026, respectively, advancing neurosymbolic multi-agent causal diagnostics for smart manufacturing.
  • 2026DETONATE was accepted to AAAI 2026 as a benchmark for text-to-image alignment and kernelized direct preference optimization.
  • 2026Three systems were accepted to AAAI Demo 2026: CausalPulse, AssetOpsBench-Live, and In-Situ Eval.
  • 2026Co-leading the AAAI 2026 hands-on lab, "From Inception to Productization: Lifecycle of Multimodal Agentic AI in Industry 4.0," in collaboration with IBM.
  • May 2025Started as a Neuro-symbolic AI Research Intern at Robert Bosch LLC, developing CausalPulse for anomaly detection, root cause analysis, and causal reasoning in manufacturing.
  • 2025SmartPilot received the Best Paper Award at IEEE CAI 2025.
  • 2025Published NSF-MAP at IJCAI 2025 and SmartPilot at IEEE CAI 2025 and AAMAS 2025.
  • 2025Presented tutorials/labs at AAAI 2025, AAMAS 2025, and US2TS 2025 on rare-event prediction, explainable multimodal AI, and multi-agent copilots for industrial AI.
  • 2025Co-organized the CODS 2025 Agentic AI Challenge with IBM and contributed to AssetOpsBench-Live.
  • 2025Published Time Series Foundational Models at AAAI 2025 and Pic2Prep at AAAI Demo 2025.
  • 2024Published rare-event and anomaly-prediction work at NeurIPS 2024 and ICMLA 2024.
  • 2024Received the Best Poster Award at DiscoverUSC 2024 for RI2AP: Robust and Interpretable 2D Anomaly Prediction in Assembly Pipelines.
  • 2026Rare Event Prediction Review paper was accepted to ACM Computing Surveys (IF: 28) , presenting the first ever survey done for rare event prediction
  • 2018Received the Gold Medal for Best Overall Performance in B.Sc. (Hons) Information Technology at KDU, Sri Lanka.
  • 2017Completed a Software Engineering Internship at WSO2, developing an XACML 3.0 policy development and evaluation tool.