I build AI-powered operations products at the intersection of logistics, data infrastructure, and automation — and I run my own AI lab at home to stay sharp on the infrastructure layer.
Most PMs use AI as a chatbot. I use it as infrastructure.
At ShipMonk (one of North America's largest 3PLs), I own the product strategy for B2B wholesale fulfillment — and I've built the AI and data tooling that makes the ops team faster: automated meeting prep, chargeback RCA assistants, SQL co-pilots, and a BI system that tracks performance across 10+ warehouses in real time.
At home, I run a self-hosted AI stack (slow-kratos) — Ollama for local LLM inference, Open WebUI, n8n for workflow automation, and Docker's MCP Gateway — all connected over Tailscale. This is where I experiment with models, build automation prototypes, and understand the infrastructure layer before I deploy anything at work.
The result: I can evaluate an AI tool from the model weights to the business outcome.
Personal (slow-kratos) |
Professional (repos below) | |
|---|---|---|
| LLMs | Ollama (local) — Llama, Mistral, Neural Chat | Claude (Anthropic), Gemini |
| Orchestration | n8n workflows, MCP Gateway | Python scripts, Asana API, REST |
| Interface | Open WebUI | Metabase, Google Docs |
| Networking | Tailscale mesh VPN | Snowflake, PostgreSQL |
| Purpose | Experiment, prototype, self-host | Production ops automation |
Warehouse Operations Intelligence Built the BI system that surfaces OTRS, OTS, pick/pack audit, and late order metrics weekly across a multi-node warehouse network — used by ops directors every week to run the business.
B2B Billing & Revenue Systems Identified and closed a structural gap in storage fee billing for wholesale merchants. Cross-system investigation spanning Snowflake, PostgreSQL, and EDI compliance data. Built the SQL framework and Metabase tooling to surface unbilled orders at scale.
AI-Powered Internal Tooling Automated the weekly ops meeting prep (2 hours → 5 minutes), built a chargeback RCA assistant that synthesizes data from 3 systems, and created a plain-English SQL co-pilot for the ops team — all using the Anthropic API.
Merchant Onboarding Analytics Python automation on the Asana API to extract and analyze go-live data across 200+ wholesale merchants — surfacing patterns in tier, warehouse assignment, integration type, and onboarding velocity.
| Layer | Personal Stack | Professional Stack |
|---|---|---|
| LLM Inference | Ollama (local GPU) | Claude API, Gemini API |
| Workflow Automation | n8n | Python, REST APIs |
| Context Protocol | Docker MCP Gateway | Anthropic MCP, Claude Code |
| Data Warehousing | — | Snowflake, PostgreSQL |
| BI & Analytics | — | Metabase (advanced SQL) |
| Networking | Tailscale | — |
| Containerization | Docker Compose | — |
| Project & Ops | — | Asana API |
| Repo | What It Is |
|---|---|
llm-daily-drivers |
Curated collection of 17 ready-to-run LLM apps for productivity, creativity, finance, health & wellness — organized by use case for non-technical users |
multi-agent-merchant-education-pipeline |
AI-powered pipeline automating merchant onboarding education with multi-agent orchestration, content personalization, and progress tracking |
| Repo | What It Is |
|---|---|
slow-kratos |
Self-hosted AI stack: Ollama + Open WebUI + n8n + MCP Gateway + Tailscale |
| Repo | What It Shows |
|---|---|
warehouse-ops-intelligence |
BI system design for B2B ops metrics at scale |
b2b-chargeback-rca-framework |
Structured RCA across a multi-system data investigation |
merchant-onboarding-analytics |
Python + Asana API automation for go-live tracking |
metabase-sql-playbook |
Documented SQL patterns for warehouse operations |
pm-data-stack-templates |
PRD, RCA, and ops meeting frameworks I use in practice |
b2b-packing-config-generator |
Google Apps Script tool for generating B2B packing configs |
opportunity-tree-framework |
Python tool converting Opportunity Tree diagrams to Excel matrices |
- Scaling AI-assisted operations tooling across multi-warehouse networks
- Building curated, user-friendly AI applications for everyday use
- Automating merchant education and onboarding workflows with multi-agent systems
- Closing revenue integrity gaps in complex B2B billing systems
- Understanding the full stack from model inference to business outcome
Open to Director of PM roles in logistics tech, data infrastructure, and AI-powered operations.
📬 LinkedIn · GitHub · Portfolio