Peer-reviewed papers and major-trade-publication articles.
Peer-reviewed papers
10 papers
Published2026
PatternForge: A Multi-Agent Generative AI Framework for Catalog-Driven Orchestration
ECAI 2026 — 18th International Conference on Electronics, Computers and Artificial Intelligence · IEEE Xplore
IEEE peer-reviewed conference. Published in IEEE Xplore (Aug 2026).
Shikher Goel and co-authors
ECAI 2026 · Submission #6585
Proposes a catalog-driven orchestration framework for multi-agent generative AI systems. Treats reusable orchestration patterns as first-class artefacts that downstream agents instantiate, compose, and reason about under enterprise governance.
PolicyAgent: A Multi-Agent Generative AI Framework for Automated Reasoning Checks
ECAI 2026 — 18th International Conference on Electronics, Computers and Artificial Intelligence · IEEE Xplore
IEEE peer-reviewed conference. Published in IEEE Xplore (Aug 2026).
Shikher Goel and co-authors
ECAI 2026 · Submission #4495
Introduces a policy-aware multi-agent layer that performs automated reasoning checks over enterprise AI workflows. Agents evaluate proposed actions against codified policy, surface violations, and provide auditable rationale before downstream execution.
Extends generative-AI agents from reactive recovery into predictive remediation and autonomous security enforcement for Infrastructure-as-Code. Proposes a resilience framework that anticipates configuration drift, surfaces security regressions before exploitation, and applies guarded remediations end-to-end across multi-cloud deployments.
Generative AIInfrastructure as CodePredictive remediationAutonomous securityCloud resilience
Introduces an autonomous-agent architecture in which Large Language Model agents continuously monitor Infrastructure-as-Code deployments, detect configuration drift and security misconfigurations in real time, and execute scoped remediations under explicit governance gates. Evaluates the framework on enterprise-cloud workloads where drift and misconfiguration form a leading source of production incidents.
LLM agentsInfrastructure as CodeConfiguration driftCloud securitySelf-healing systems
Defines a multi-agent AI framework that detects deployment-failure precursors across multi-cloud environments and applies autonomous remediations under bounded policy. Presented at ECAI 2026, Bucharest/Pitești (Jul 2–3, 2026); proceedings published in IEEE Xplore.
Applies zero-trust principles to autonomous Infrastructure-as-Code security remediation. A multi-agent architecture verifies every identity, action, and remediation against an explicit trust boundary before execution — translating zero-trust theory into operational IaC controls. Published in IEEE Xplore as part of the ECAI 2026 proceedings.
Multi-agent AIZero trustInfrastructure as CodeAutonomous remediationCloud security
IEEE peer-reviewed conference (Cloud Summit 2026), Washington, DC, USA, 25–26 June 2026. Added to IEEE Xplore on 31 July 2026. 41 full papers and 20 posters accepted from 119 submissions. Publisher: IEEE.
Running cloud-native systems in production requires more than deploying containers into Kubernetes. The platform must provide repeatable infrastructure provisioning, controlled release workflows, safe traffic migration, operational visibility, and practical integration with existing enterprise systems. This paper presents an industry case study of a production-scale platform on Amazon Web Services (AWS) built around Amazon Elastic Kubernetes Service (EKS). The implementation combines Terraform for infrastructure provisioning, GitLab CI for build automation, ArgoCD for GitOps-based delivery, and a blue-green release model using separate active and passive Kubernetes environments. Traffic is introduced to the passive environment incrementally after validation checks, with routing controlled through Route53, AWS Application Load Balancer (ALB), and Istio ingress policies. Runtime elasticity is supported through KEDA for workload-driven pod scaling and Karpenter for dynamic node provisioning. The platform also integrates Amazon ElastiCache for Redis as a shared low-latency data layer and AWS Direct Connect for hybrid communication with on-premise services. The paper focuses on the operational design, deployment process, traffic switching strategy, validation gates, and lessons learned from applying these patterns to real production operations.
Presents a reference model for embedding agentic AI inside enterprise customer-relationship platforms. Maps a multi-agent architecture onto Salesforce CRM and AWS data-and-integration infrastructure, addressing identity, tool access, governance, and audit boundaries required for regulated deployment.
Compliance-Driven Security Automation in AI-Augmented CI/CD Pipelines
SCI 2026 — 8th International Conference on Smart Computing & Informatics · Springer LNNS
International peer-reviewed conference, proceedings indexed in Springer LNNS and Scopus. Recognised with the Best Paper award (per author notification).
Synthesises CI/CD pipeline architectures for production enterprise platforms operating under SOX, FINRA, and SEC controls. Embeds AI-augmented stages for segregation of duties, automated control-evidence collection, and audit-traceable change management. Lessons drawn from large-scale Salesforce and AWS deployments inside a Global Systemically Important Bank.
A Pattern Catalog for Multi-Agent Generative AI Orchestration Across Enterprise CRM and Hyperscaler Cloud Platforms
IEEE Access — Journal
IEEE peer-reviewed open-access journal (impact-factor indexed). Manuscript Access-2026-28862 submitted 14 Jun 2026; under associate-editor review.
Shikher Goel and co-authors
IEEE Access · MS Access-2026-28862
Curates a pattern catalog for multi-agent generative-AI orchestration spanning enterprise CRM platforms (Salesforce) and hyperscaler cloud (AWS, Azure). Treats each pattern as a reusable, governed unit that practising architects can compose into production systems.
Forbes Technology Council is an invitation-only community of senior technology leaders; Council articles are editorially reviewed by Forbes staff. Shikher was a Council Member during 2025 and authored the pieces below in that capacity; the articles remain published on Forbes.com.
View contributor profile →
This page surfaces only formal scholarly venues and major-trade publications. Long-form essays, walkthroughs, and Medium posts are catalogued separately on the Writing page.