Defining the security discipline for AI agents, autonomous systems, and agentic workflows.
25+ years of enterprise architecture. Now applied to the hardest problem in AI.
Current Project
Building AISecOps — the open security specification and runtime interceptor for agentic AI systems.
IT Architect and Senior Software Engineer with 25+ years delivering secure, scalable enterprise applications for Fortune 1000 companies and government agencies. Expert in Java, Spring Boot, cloud-native platforms (AWS, GCP, Azure), and DevSecOps with deep security tooling integration across CI/CD pipelines.
Recent focus on AI-powered solutions using LangChain, RAG, Ollama, OpenAI API, PyTorch, and TensorFlow — building intelligent automation, real-time analytics, and self-hosted LLM workflows. Skilled in threat modeling, AI threat review, and secure SDLC for high-compliance environments.
Proven leader in Agile and Scrum environments, mentoring teams and modernizing legacy systems with cutting-edge technology.
With 25+ years of enterprise delivery, I favor pragmatic architecture choices over trend chasing.
I build durable platforms with clear trade-offs, biasing for security, observability, and performance at scale.
Languages
Java (8-21), TypeScript, Python
Frameworks
Spring Boot, React, FastAPI
Cloud
AWS, GCP, Azure, Kubernetes
Data
PostgreSQL, BigQuery, Cassandra
AI/ML
LangChain, RAG, PyTorch, Ollama
DevSecOps
GitHub Actions, Terraform, SAST/DAST
Self-Employed
The Home Depot SFO
The Home Depot Warehouse
CVS Health / Aetna
The Home Depot (Cart Commons)
The Home Depot (Finance IT)
Avaya & Lucent Technologies
Production-ready AI/ML solutions and innovative prototypes demonstrating modern engineering practices and security-first architecture.
Hands-on security lab for agentic AI systems featuring runtime guardrails, policy enforcement, sandboxed tool execution, and full audit traceability.
Production-style execution gateway for tool-using agents with request mediation, policy validation, controlled tool invocation, and observability hooks.
Extends OpenClaw with secure tool runner patterns enabling policy-checked execution, sandboxing, and traceable outcomes for enterprise AI workflows.
Comprehensive STRIDE-based threat model for multi-agent systems covering identity, memory isolation, tool boundaries, and data exfiltration mitigation.
Cross-modal Retrieval-Augmented Generation platform supporting documents, audio, and video with transcription, embeddings, and semantic search via interactive UI.
End-to-end trading system featuring backtesting, live data streaming, predictive modeling with PyTorch, and dashboard-driven strategy orchestration.
Secure edge-based IoT controller built on Raspberry Pi demonstrating authenticated device orchestration, command validation, and safe physical actuation patterns.
Automated STRIDE-based threat modeling embedded into CI pipelines. Generates structured Markdown risk reports directly from pull request diffs using Large Language Models.
Natural-language powered log investigation system integrating LangChain with local Large Language Models and Elastic stack for accelerated root cause analysis.
Medium deep-dives, talks, and field-notes on shipping secure agents, resilient APIs, and systems that actually run in production.
Open to opportunities in: