Director of Engineering | AI Innovator
Citi - Capital Markets Technology
Building the future of AI-powered enterprise systems. Leading cloud-native platform engineering, pioneering chaos engineering practices, and creating open-source tools that bridge AI with enterprise Java.
Building autonomous AI agents
Kubernetes & OpenShift
Resilience & testing
LLM + Neo4j integration
With over 20 years in software development and technology leadership, I specialize in designing AI-powered cloud-native platforms for low-latency financial applications at Citi. My work spans DevOps automation, chaos engineering, and the development of intelligent systems that enhance enterprise resilience and efficiency.
I'm passionate about bridging the gap between cutting-edge AI research and enterprise Java applications. My open-source project Tools4AI demonstrates this vision, providing a 100% Java framework for building autonomous AI agents that integrate seamlessly with existing enterprise infrastructure.
Through technical writing on Medium and LinkedIn, I actively contribute to the AI community by sharing practical knowledge on LLMs, agentic AI, and enterprise Java integration strategies.
AI compliance, chaos testing, ML systems
111 repositories, 350+ GitHub stars
Medium & LinkedIn publications
Pioneering solutions in AI governance, chaos engineering, and cloud resilience
US-12,621,253
May 5, 2026
Automatically evaluate, select, and coordinate AI-based agents for collaborative distributed task execution based on dynamic multi-attribute scoring and resource allocation models.
View PatentUS-12,602,418
April 14, 2026
Decomposes queries into sub-queries, routes to specialized models, detects conflicts among outputs, and generates aggregated responses with conflict resolution.
View PatentUS-12,596,738
April 7, 2026
Selects LLMs based on complexity, domain, and compliance parameters while generating human-readable explanations and tamper-evident audit trails.
View PatentUS-12,592,897
March 31, 2026
Automatically detect, analyze, and mitigate anomalous resource distribution among AI agents within distributed computational networks.
View PatentUS-12,587,490
March 24, 2026
Automatically authorize, audit, and manage usage of protected digital content via agentic AI models with cryptographic signatures and audit trails.
View PatentUS-12,587,489
March 24, 2026
Automatically register, monitor, and authenticate distributed AI agents using a distributed ledger-based agent knowledge registry.
View PatentUS-12,517,724
January 6, 2026
Uses AI models to identify gaps in system assets and generate actions to ensure observed assets meet identified criteria.
View PatentUS-12,483,471
November 25, 2025
Dynamically monitors document repositories and updates communication maps to detect and remediate validation failures for network operations.
View PatentUS-12,450,494
October 21, 2025
Evaluates autonomous agents by identifying gaps in proposed actions and modifying them using generative AI models.
View PatentUS-12,361,335
July 15, 2025
Employs specialized models to identify non-compliance with vector constraints within AI-generated responses.
View PatentUS-12,299,140
May 13, 2025
Multi-model superstructure that dynamically routes artifacts for evaluation and generates corrective actions when metrics are not met.
View PatentUS-12,198,030
January 14, 2025
Real-time detection and automatic correction of bias, harmful content, and IP violations in AI outputs.
View PatentUS-12,184,480
December 31, 2024
Dynamic monitoring, evaluation, and mitigation of detected anomalies in real-time for conglomerate-application ecosystems.
View PatentUS-12,111,754
October 8, 2024
Constructs test cases with prompts and expected outcomes to evaluate AI application compliance with guidelines.
View PatentUS-11,924,027
March 5, 2024
Reduces wasted computational resources by detecting and preventing invalid network operations from propagating.
View PatentUS-11,847,046
December 19, 2023
A single network system providing functionality, resiliency, chaos, and performance testing for cloud applications.
View PatentBuilding bridges between AI and enterprise Java ecosystems
100% Java-based Agentic AI Framework for building autonomous agents. Converts natural language to HTTP REST calls, Java methods, shell scripts, and Swagger API calls. Supports multi-AI voting, hallucination detection, and action risk assessment.
View ProjectPure Java implementation of Google's A2A (Agent-to-Agent) protocol for Spring Boot. Agents automatically exposed as MCP tools.
View ProjectArduino/ESP32 robotics integration with Claude AI via MCP protocol. Enables AI-driven hardware control.
View ProjectGit-based federated knowledge system extension for BMAD-METHOD.
View ProjectSharing insights on AI, LLMs, and enterprise integration
A comprehensive guide to integrating knowledge graphs with LLMs for improved accuracy and grounding.
Read ArticleBuild your own LLM-powered alternative to Drools with this comprehensive Java guide.
Read ArticleExploring the intersection of knowledge graphs and Model Context Protocol for enhanced AI capabilities.
Read ArticleStep-by-step guide to writing your first AI code with TinyLlama and understanding LLM fundamentals.
Read ArticlePractical patterns for integrating autonomous AI agents into enterprise Java applications.
Read ArticleGuide to quantizing models and optimizing performance for consumer-grade CPUs.
Read ArticleAsk me anything about my background, patents, or expertise. The AI runs entirely in your browser — no data leaves your device.
Runs locally via WebLLM + WebGPU. First load downloads model weights (~1–2 GB, cached in browser after).
Interested in collaboration, speaking opportunities, or discussing AI innovations?