Engagement Manager & Enterprise Solution Architect at Cybage, hands-on with the code. Databases in 2008, and I kept following the hard part upward: distributed systems, multi-cloud, and now the AI layer.
A bit of what I work with, hands-on:
- 🛠️ Automation
Goand Python tooling, Boto3 monitoring, in-house scripts against cloud DevOps APIs, IaC pipelines. If a task repeats, I script it away. - 🤖 AI writing
MCP servers, governed multi-agent delivery systems, and agentic workflows; building gold-tier data for AI/ML. - 🗄️ Databases zero-downtime migrations across
OracleMSSQLMySQLPostgreSQLCassandra; HA/DR, tuning, and VLDB work hands-on. - ☁️ Cloud
AWS·GCP·OCI, data platforms onDatabricks·Redshift·BigQuery·Snowflake·Kafka.
I like staying close to the build; the architecture I hand a team is usually something I've prototyped first. Most of what's here is built after hours, a night-time builder's habit. And when I step away, my agents keep going, working through the tasks I've handed them while I'm elsewhere.
Sometimes I write it down → Medium · dev.to
📍 Minneapolis–St. Paul
| Project | What it is |
|---|---|
| specter-agent | Governed multi-agent delivery workflows: design, run, approve, audit from one workspace. FastAPI + React. |
| Relayent | Job broker that runs AI jobs on a machine's CLI subscription. Multi-tenant relay + bridge, in Go. |
| nj-agents | SDLC AI-agent toolkit: skills + agents across the whole software lifecycle. |
| mcp-server-salesforce | MCP server for AI assistants to work with Salesforce orgs: Apex, SOQL, metadata. |



