The deployment layer. I embed in client environments and make agents work against data that was never prepared for them.
You wrote a job for the person who makes the promise real. That is the only work I have done for twenty years — Informatica’s parachute closer, then the same playbook on a private RTX 4070 agent stack that measures. Clients asked for me by name. I go in alone and come back with something that runs.
OpenClaw WhatsApp agents, 10+ skills. Multi-agent pipeline (systems → architect → validator) on local Ollama. Dockroot-MCP: agents see Docker/network, not host-root.
World-first Informatica Grid/HA + DR, 20+ nodes, four regions. local_grid_suite: SQLite benches, GPU routing, CLI + PDF. Measured ~27× decode vs CPU baseline.
Ameriprise IDQ / EDC / Axon. Type 2 warehouse at Dawn. FDA-validated platform at Smith & Nephew. Vocabulary in the tables vs vocabulary in the heads.
Containerized private LLM on RTX 4070 / WSL2: TabbyAPI EXL3, Ollama, Open WebUI, CUDA, flash-attention, q8 KV. huggingface.co/jpanasuk/tabby-tavern-stack
Model discovery, decode/prefill benches, GPU routing, SQLite logs. Checked-in: 1.39 → 37.47 tok/s after routing. Warm stream ~76 tok/s.
Bounded agent visibility into Docker and network wiring. Same Grid/HA instinct: measure the boundary, then hand the agent the boundary — not the box.
Production WhatsApp agents. Structured output, repair retries, collision checks. Multi-provider routing with cloud failover so one surface covers local and hosted models.
Also retained through 2025 for Informatica production rescue: Grid/HA, PowerExchange CDC, IDQ, v9/v10, SOAP/REST, SOW/sizing, install-to-go-live. Available remote. High-autonomy teams only.
I just came off two years of laydown. Not idle — a rebuild. I stood up a private GPU lab, measured it, published it, and set one goal: use the new stack to help land the AI revolution in rooms that still run on unprepared data.
Off the corporate conveyor on purpose. Two years to retool: agents, bounded MCP, local inference that reports numbers, not slides. The same diagnostic playbook as Informatica, now aimed at Forge / Lattice / Stratum.
Remote from Wisconsin. No dual-career relocation theater. High autonomy. I can embed in a client environment, stay through production, and still be the person who comes back with a playbook.
You need someone who can go in alone now, while the enterprise motion is being built. I am available, current, and not splitting attention with a job I am trying to leave. This is the fit.
Talend 8 mappings on a live platform. GitLab CI/CD migrations. Redshift + SQL Server. 99.99% stability. On-call incident command and root-cause writeups.
PepsiCo Grid/HA and Data Integration Hub across NA/LATAM. Led v9→v10 with zero data loss. Teradata → Big Data (~40% analytics lift). Ameriprise EDG: IDQ, EDC, Axon, 24×7. Ace Hardware PowerCenter → Big Data objects with business + QA.
Analyst → Consulting Manager. Teams of 7–14. $4M+ services in two years. Consultant of the Year runner-up (2016). Parachute closer: 20 troubled accounts entered, 19 recovered. Playbooks so the next one went faster.
Medicare + MTM data marts. Delivered above grade-level. Client: trusted to ship with quality.
24/7 data integration and warehousing. CFO Award 2005 (six awarded per year). Quarterly Award for Excellence 2006, Demantra.
Class B fireworks shooter. Inventory manager, $800K+ seasonal stock. No schema, no second take. The line: customer satisfaction is worthless; customer loyalty is priceless.
AI / agents · Python · Docker · CUDA · Ollama · TabbyAPI EXL3 · MCP · RAG · OpenClaw · SQLite telemetry · Linux/WSL2
Data · SQL · Informatica PowerCenter · Grid/HA · PowerExchange CDC · IDQ · EDC · Axon · DIH · Talend 8 · Oracle · SQL Server · DB2 · Teradata · Redshift · GitLab CI/CD
Domain · Banking · insurance · wealth · healthcare/PBM · CPG · manufacturing · FDA-validated platforms
Education · BBA MIS, University of Wisconsin–Eau Claire · Informatica Certified Specialist (Administrator, Big Data Administrator, PowerCenter Developer I & II)