Building secure infrastructure • Reliable platforms • Distributed systems • Frontier AI evaluation environments
I am a DevSecOps Engineer, Site Reliability Engineer, Cloud & Platform Engineer, and AI Benchmark Engineer focused on secure production systems and hard technical evaluation environments.
My work spans Kubernetes, cloud infrastructure, CI/CD, GitOps, observability, distributed systems, Hyperledger Fabric, microservices, RAG infrastructure, application security, and frontier AI evaluation.
Security + Reliability + Automation + Distributed Systems + AI Evaluation
- ☸️ Kubernetes and cloud platform engineering
- 🔐 DevSecOps, AppSec, IAM, and software supply-chain security
- 🔄 CI/CD, GitOps, release engineering, and Infrastructure as Code
- 📊 Prometheus, Grafana, distributed tracing, and operational alerting
- ⛓️ Hyperledger Fabric, consensus, ledger, identity, and chaincode infrastructure
- 🤖 Terminal-based agent benchmarks and coding-agent evaluation
- 🧪 Oracle, verifier, rubric, regression, and failure-analysis engineering
- 🧠 RAG gateways, multi-provider inference, and AI infrastructure
Related infrastructure:
My merged TreeTracker contribution integrates Keycloak authentication with Hyperledger Fabric identity operations.
Keycloak • JWT authentication • Fabric CA • wallet management • API endpoints • Docker • security configuration
Kubernetes-based Hyperledger Fabric infrastructure covering certificate authorities, peers, orderers, CouchDB, channels, connection profiles, and chaincode lifecycle.
Grafana = Fabric health and performance
Prometheus = metrics collection
Alertmanager = operational alerts
Hyperledger Explorer = blockchain visibility
Related infrastructure:
Community:
Productionization work across the TreeTracker and Hyperledger Fabric stack:
- TreeTracker web, admin, authentication, capture, and token services
- Kubernetes and Argo CD GitOps
- GHCR image delivery
- PostgreSQL and Keycloak migration and persistence
- Fabric peers, orderers, CAs, CouchDB, and wallets
- Hyperledger Explorer and CouchDB Fauxton
- Headlamp, Prometheus, and Grafana
- Backup, restoration, credentials, and persistence documentation
Related delivery tooling:
My TreeTracker Wallet work covers GitHub Actions CI/CD, Docker builds, security scanning, SBOM generation, and Kustomize deployment overlays.
My Prometheus, Grafana, Loki and Alertmanager deployment repositories are derived from my running Hyperledger Fabric / TreeTracker stack. SigNoz is an additional deployment package. Each includes Helm charts, native Kubernetes manifests, Terraform and OpenTofu deployments, high-level architecture, and operations runbooks.
Prometheus, Grafana, Loki and Alertmanager were validated in an isolated cluster: metrics collection, 15 Grafana dashboards, log ingestion and queries, alert handling, and persistence across pod restarts. Production profiles document their storage, identity, availability, and recovery requirements.
Evidence-backed incident reports covering impact, root cause, recovery, verification, and follow-up reliability work.
| Incident date | Postmortem | Incident summary | Dated tag |
|---|---|---|---|
| 2026-09-19 | Windows/WSL Platform Recovery and Terminal-Bench Runner Offline | Restored the paired runner and automatic platform recovery; an actual reboot verified accepted heartbeats before Windows sign-in. | incident-2026-09-19 |
| 2026-09-16 | Fixam4me Booking Service Restart Incident | Repeated restarts followed an unhandled database query rejection; the dependency connectivity cause and durable fix remained open. | — |
| 2026-09-08 | Hyperledger Fabric Development Outage — SRE Postmortem | Recovered eight CouchDBs, eight Fabric peers, and Explorer after Erlang startup OOM and identity/TLS trust failures. | incident-2026-09-08 |
My evaluation-engineering capabilities span secure systems, software engineering, infrastructure, distributed systems, and AI-agent evaluation.
- Containerized execution environments and reproducible test workflows
- Reference solutions, acceptance criteria, and independent verifiers
- Regression testing, edge-case analysis, and failure diagnosis
- Repository investigation and coding-agent evaluation
- Application security, infrastructure reliability, and distributed-systems testing
Harbor and Terminal-Bench are public upstream projects, credited to their respective maintainers and contributors.
ExtractBench is a public schema-guided document-extraction benchmark from run-llama and its contributors, released under the Apache License 2.0. My repository is a fork of that upstream project.
My component deployment repositories below include Kubernetes manifests, Helm, Terraform, OpenTofu, architecture diagrams and operations runbooks. Database packages use three-member HA topologies; each repository records its validation evidence and remaining production acceptance requirements.
Focus: AppSec • IAM • OAuth/OIDC • JWT • DPoP • mTLS • container security • secret detection • CI/CD security • Kubernetes security • TLS • secure software delivery • runtime protection • supply-chain security
Application and Kubernetes/platform engineering for a service marketplace platform.
Security audit work included CLI transfer and cold-signing attack surfaces.
kind-dev maintenance work included zero-ready ReplicaSet cleanup and cluster troubleshooting.
Windows/WSL performance and disk-storage diagnostics.
- Production-grade Hyperledger Fabric on Kubernetes
- Automated cryptographic material and certificate lifecycle management
- Fabric observability with Prometheus, Grafana, and Alertmanager
- Kubernetes platform engineering and GitOps
- Secure CI/CD and AppSec automation
- AI-agent evaluation and benchmark engineering
- RAG infrastructure and multi-provider inference
- Cloud-native microservices and distributed systems
I am interested in DevSecOps, SRE, cloud infrastructure, AI systems, distributed systems, security engineering, and platform engineering opportunities and collaborations.
