Research & publications
Research
Nine IEEE-published papers on reliability, evaluation, and guardrails for agentic AI, distributed systems, cloud platforms, and applied machine learning. View my complete record onIEEE Xplore ↗.
GuardGRPC: Priority-Aware Resource Governance for OTLP/gRPC Telemetry Export
IEEE · 2026 5th International Conference on Electronics Representation and Algorithm (ICERA), Yogyakarta, Indonesia
A priority-aware governance layer for OTLP/gRPC export that adapts telemetry controls under pressure while preserving incident-critical signals.
Kairos: SLO-Aware Scheduling and a Benchmark for Multi-Tenant LLM Inference Serving under Bursty, Heterogeneous Load
IEEE · 2026 5th International Conference on Electronics Representation and Algorithm (ICERA), Yogyakarta, Indonesia
An SLO-aware scheduling approach and reproducible benchmark for multi-tenant LLM inference under bursty, heterogeneous workloads.
Priority-Aware Edge-to-Cloud IoT Event Streaming for Reliable Proof-of-Delivery in Mission-Critical Logistics
IEEE · 2026 5th International Conference on Electronics Representation and Algorithm (ICERA), Yogyakarta, Indonesia
A priority-aware edge-to-cloud event-streaming design for preserving reliable proof-of-delivery evidence in mission-critical logistics.
Containing the Cascade: A Benchmark and Reference Mediator for Failure Propagation in Tool-Using LLM Multi-Agent Systems
IEEE · 2026 5th International Conference on Electronics Representation and Algorithm (ICERA), Yogyakarta, Indonesia
A taxonomy of cascading failures, the Cascade-Bench fault-injection testbed, and Sentinel, a reference mediator for containing failures in tool-using multi-agent systems.
GuardedProcure: Validator-Guided Typed Planning for Reliable B2B E-Commerce Interoperability
IEEE · 2026 5th International Conference on Electronics Representation and Algorithm (ICERA), Yogyakarta, Indonesia
A deterministic replay benchmark and validator-guided typed planner for reliable procurement-message transformation under schema drift.
Human-Assured Agentic Interoperability for Resilient B2B Systems: A Deterministic Replay Benchmark for Procurement Message Transformation
IEEE · 2026 5th International Conference on Electronics Representation and Algorithm (ICERA), Yogyakarta, Indonesia
A deterministic replay benchmark for B2B procurement transformation that combines risk, novelty, and reviewer readiness to govern autonomous actions.
SecReviewAgent: Context-Aware Security Review of Infrastructure-as-Code Using Persistent Architecture Memory
IEEE · 2026 Seventeenth International Conference on Ubiquitous and Future Networks (ICUFN), Milan, Italy
An LLM-powered Infrastructure-as-Code security-review system that uses persistent architecture memory to evaluate changes in repository context.
HybridFraudNet: A Dual-Branch CNN-GNN Architecture for Healthcare Insurance Fraud Detection
IEEE · 2026 Seventeenth International Conference on Ubiquitous and Future Networks (ICUFN), Milan, Italy
A dual-branch architecture that fuses claim-level CNN signals with provider-network GraphSAGE embeddings for healthcare insurance fraud detection.
CostAgent: Self-Improving Autonomous LLM-Based Orchestration for Cost-Optimal Cloud Data Processing at Scale
IEEE · 2026 International Conference on Artificial Intelligence, Systems, and Emerging Technologies (ICAISET), Cairo, Egypt
A bounded-risk autonomous orchestration framework for cost-aware cloud data processing on preemptible infrastructure with deterministic safety enforcement.