Tom Pinder
Operations and systems engineer who deploys AI with guardrails.
I spent 15 years running warehouse and supply-chain systems. Now I deploy AI into production with a governance and accuracy focus. The rare part is the overlap: an operations person who has run a distribution floor and also builds the guardrails that keep AI honest.
What I work on
AI guardrails and safe deployment
I build the layer that stops AI from shipping wrong answers. For PRAPI I built the anti-fabrication system: a claim-gate that blocks unverifiable quantitative claims before anything publishes, a send-guard that refuses unresolved placeholders, and grounding that ties generated output back to a verifiable source. I authored the brief.md spec, an open format for giving an LLM the right inputs so it stops hallucinating. The throughline is integrity: AI systems that refuse instead of guess.
Warehouse and supply-chain operations
I ran Manhattan SCALE WMS across six distribution centers. I built operational reporting on SQL Server and T-SQL, with 10 years in SSRS and shorter stints in Snowflake and Cognos, in Azure. The work was the real core of distribution: WMS rollouts and go-lives, inventory accuracy, throughput and bottleneck diagnosis, slotting, and the reports operators actually use to run a floor.
The intersection
Having run a distribution center's systems, I know where AI helps, where it is theater, and where letting a model touch an inventory decision without a guardrail is a real risk. That is the forward-deployed-engineer-in-logistics profile: translate between the people who run operations and the AI that is supposed to help them, and ship it without breaking trust.
Background
- 15 years in operations and systems engineering.
- Manhattan SCALE WMS across six distribution centers.
- SQL Server and T-SQL; SSRS (10 years); Snowflake and Cognos; Azure.
- Built PRAPI's accuracy and anti-fabrication layer. Authored the brief.md spec (theintegrityframework.org).
- USMC combat veteran (0311 Infantry / RTO). Held a Secret clearance about 20 years ago; it is lapsed and not active.
- Based in Hampstead, NC. Remote.
Currently
Open to remote roles where operations depth and correct AI deployment meet: forward-deployed and AI-deployment engineering, implementation and solutions consulting, BI and SQL development, data and supply-chain analysis, and WMS.
FAQ
- What does Tom Pinder do?
- He is an operations and systems engineer who deploys AI into production with accuracy guardrails. He has 15 years in warehouse and supply-chain systems and builds the anti-fabrication layer that keeps AI output verifiable.
- What is Tom Pinder's AI-deployment background?
- He built PRAPI's claim-gate, send-guard, and grounding layer, which block fabricated or unverifiable output before it ships, and he authored the brief.md spec for portable context. His focus is deployment and governance, not model training.
- What is Tom Pinder's supply-chain and warehouse experience?
- He ran Manhattan SCALE WMS across six distribution centers and built operational reporting on SQL Server, T-SQL, SSRS, Snowflake, and Cognos, covering WMS rollouts, inventory accuracy, throughput, and slotting.
- Is Tom Pinder available for work?
- He is open to remote roles in forward-deployed and AI-deployment engineering, implementation and solutions consulting, BI and SQL development, data and supply-chain analysis, and WMS.