Jayapragash Dakshnamurthy
IEEE Senior Member, Senior Software Engineer, and Independent Researcher with 18+ years of experience designing cloud-native platforms, distributed systems, microservices, production AI systems, and resilient enterprise software.
About
I specialize in cloud architecture, distributed systems, microservices modernization, observability, cybersecurity, software reliability, and AI-assisted engineering. My work focuses on designing scalable and dependable platforms, improving production performance, and modernizing complex enterprise systems.
Alongside product engineering, I contribute through technical publications, peer-reviewed research, conference Program Committee service, and academic reviewing.
My research interests include reliable AI systems, secure agentic execution, policy-driven runtime governance, failure containment, latency amplification, graph-based anomaly detection, root-cause localization, neural-network reliability, and cyber-resilient distributed architectures.
Current Highlights
IEEE Senior Member
Elevated to the grade of IEEE Senior Member in September 2026.
IEEE Xplore Publication
Peer-reviewed SmartCloud 2026 research published and indexed in IEEE Xplore.
Accepted Research Paper
Weight-Only Spectral Detection of Neural Network Overfitting Using Random Matrix Theory accepted for IC-SQITS 2026.
Accepted · Awaiting Indexing
Policy-Driven Runtime Governance for Secure and Trustworthy Enterprise Agentic AI accepted for AISRI 2026; proceedings indexing is pending.
Forthcoming Book
Author of Building Reliable AI Systems, focused on dependable production AI architecture and operations.
Published DZone Articles
Published engineering work covering scalability, caching, distributed architectures, and enterprise modernization.
DZone Article Views
More than 4,000 combined views across currently published DZone technical articles.
Program Committee Roles
Program Committee service across MIWAI 2026, IC-SQITS 2026, and IEEE BigData 2026.
Conference Reviewer Roles
Peer-review contributions across AI, cybersecurity, distributed systems, cloud computing, and intelligent systems.
Featured IEEE Publication
The Hidden Cost of AI Systems: Latency Amplification, Retries, and Cascading Failures in Production
Peer-reviewed research examining latency amplification, retry behavior, cascading failures, and resilience challenges in production AI and distributed systems.
Author: Jayapragash Dakshnamurthy
Publisher: IEEE
Year: 2026
IEEE Xplore Document: 11638708
Featured Technical Writing
Why Push-Based Systems Fail at Scale — and How Hybrid Fan-Out Fixes It
Explains why large fan-out workloads create operational pressure and how hybrid delivery architectures improve scalability.
Every Cache Miss Is a Tiny Tax on Your Performance
Examines how cache misses increase latency, backend load, and infrastructure cost in high-scale systems.
From Monolith to Microservices: Practical Lessons From Real System Modernization
Practical lessons involving architectural boundaries, observability, DevOps, data ownership, and migration risk.
Reliable AI Systems
Engineering writing on production AI reliability, observability, agentic architectures, distributed systems, failure modes, and enterprise AI adoption.
Book
Building Reliable AI Systems
A practical engineering book focused on designing, deploying, operating, securing, and governing reliable AI systems in enterprise environments.
Key topics include:
- Production AI reliability
- Continuous evaluation
- AI observability
- Agentic AI architecture
- Security and runtime governance
- Human oversight
- Multi-agent systems
- AI FinOps
- Operational resilience
Research & Conference Publications
Published Research
The Hidden Cost of AI Systems: Latency Amplification, Retries, and Cascading Failures in Production
Research examining latency amplification, retry storms, cascading failures, and resilience patterns in production AI and distributed systems.
IEEE XploreAccepted Research
Policy-Driven Runtime Governance for Secure and Trustworthy Enterprise Agentic AI
Research presenting a policy-driven runtime governance architecture for secure and trustworthy enterprise agentic AI, with emphasis on runtime policy enforcement, governance controls, human oversight, auditability, and secure agent execution.
Status: Conditionally accepted for presentation and publication. Final acceptance is subject to successful evaluation of the revised manuscript.
Accepted Conference Research
Weight-Only Spectral Detection of Neural Network Overfitting Using Random Matrix Theory
Accepted for publication in the conference proceedings. The work investigates weight-only spectral indicators for detecting neural-network overfitting using Random Matrix Theory.
Toward Trustworthy Enterprise AI Agents: A Human-Governed Runtime Architecture for Secure Execution
Research on human-governed runtime architecture for trustworthy enterprise AI agents, emphasizing policy enforcement, controlled execution, security, and human oversight.
Design and Implementation of Secure Agentic AI Systems Using Token Vault Architecture for Real-Time Enterprise Applications
Research focused on secure agentic execution, token vault architecture, credential isolation, controlled access, and auditability.
Active Research Submissions
Observability-Driven Failure Containment in Distributed Systems: An Experimental Study of Retry, Circuit Breakers, and Latency Amplification
Research studying retry amplification, circuit-breaker behavior, latency propagation, and observability-driven failure containment.
Cyber-Resilient Edge-Cloud Microservices: AI-Driven Decentralized Anomaly Detection and Root Cause Localization Using Graph Neural Networks
Research on decentralized anomaly detection and root-cause localization in edge-cloud microservices using dependency-aware Graph Neural Networks.
Professional Service, Reviewer Roles & Technical Program Committees
Cyber-AI 2026
Served as Session Co-Chair, supporting session coordination and technical program delivery.
IEEE Access
Reviewer service covering four IEEE Access manuscripts.
ICAIC 2027
Reviewer supporting technical evaluation of AI and cybersecurity research submissions.
AISRI 2026
Completed assigned peer-review service for AISRI 2026.
IEEE BigData 2026
Serving as a Program Committee member for the IEEE International Conference on Big Data 2026, supporting peer review and technical evaluation of research submissions.
MIWAI 2026
Program Committee service supporting peer review and technical evaluation of research in artificial intelligence and intelligent systems.
IC-SQITS 2026
Program Committee member for the First International Conference on Secure Quantum Intelligence and Trusted Systems.
IC-SQITS 2026
Reviewing research involving trustworthy AI, cybersecurity, adversarial machine learning, autonomous systems, and intelligent security architectures.
ICMACC 2026
Invited reviewer for the 2026 International Conference on Advances in Computing, Communication and Materials.
GAISS 2026
Peer-review service supporting evaluation of research involving generative AI, secure systems, and emerging AI architectures.
IEEE SmartCloud, AISRI, Cyber-AI, GAISS, IC-SQITS, ICMACC & ICAIC
Research contributions spanning distributed systems, cybersecurity, cloud computing, reliable AI, agentic systems, runtime governance, and enterprise architecture.
IEEE Senior Member · Princeton / Central Jersey Section
Elevated to the grade of IEEE Senior Member in September 2026, recognizing significant professional experience and technical contributions. Continuing engagement includes research, conferences, peer review, leadership, and technical communities.
Publications & Professional Profiles
- IEEE Xplore: SmartCloud 2026 Publication
- Google Scholar: Academic Research Profile
- ORCID: Research Profile
- DZone: Technical Articles
- LinkedIn: Professional Profile & Newsletter
- Substack: Engineering Articles
- Medium: Technical Publications
- GitHub: jaya12pragash
Contact & Connect
For professional collaboration, technical discussions, research, conference reviewing, Program Committee opportunities, and engineering community activities, connect through the profiles below.
LinkedIn: linkedin.com/in/jayapragashd
GitHub: github.com/jaya12pragash
Substack: jayapragashdakshnamurthy.substack.com