Supply ChainIntelligence

Peakview Labs builds working AI systems for supply chain operations. Insight is a risk-management platform that watches your network and manages the response, not just the alert. Forge builds custom supply chain software, with AI carrying the implementation so a working system arrives in weeks.

The Work

What Peakview Labs Built

Five working prototypes, each taken from problem framing through to running code and exercised against synthetic supply chain data. Every system targets a problem drawn from two decades of running supply chains at Amazon, CVS Health, and Clean Harbors.

Inventory Orchestration Agent

The problem

Routine C-items are cheap enough to ignore and expensive enough to stop a line.

The build

Watches inventory against the production schedule and places routine purchase orders on its own below a set spend threshold, holding anything above it for a person. Every action is written to an audit log next to the rule that allowed it.

PO-1042  placed  $412  rule C-07
PO-1043  placed  $268  rule C-07
PO-1044  HELD    $2,150  over threshold
Illustrative
PythonAWS LambdaClaudePostgreSQLRedis

Cold-Chain Prediction

The problem

A refrigeration failure is usually discovered after the load is already lost.

The build

Ingests refrigeration telemetry and flags temperature drift early enough to act on it, rather than reporting the loss afterwards.

Illustrative
AWS IoT CoreKinesisSageMakerPython

Document-to-EDI Pipeline

The problem

Healthcare remittance arrives as unstructured documents and has to become valid EDI.

The build

Converts EOB documents into EDI 835 transactions. The model handles document understanding; a deterministic rule validator checks every field before anything is emitted, so the model never has the last word on correctness.

ISA*00*0043
CLP*A314*04
SE*22*0001
Illustrative
ClaudePythonRule validator

Dynamic Inventory Rebalancing

The problem

Fixed min/max levels stop reflecting reality the moment demand shifts.

The build

Moves stock between locations based on observed velocity instead of static reorder points, so slow sites release inventory that fast sites need.

Illustrative
PythonAWS FargateRedis

Capacity Digital Twin

The problem

Sales and operations argue about capacity from two different spreadsheets.

The build

Simulates plant capacity against demand scenarios so both sides are arguing from one model. Built to remove the loudest-voice-wins dynamic from S&OP.

Illustrative
PythonAWS FargateNode

Peakview Insight

Full-platform build

A supplier risk-intelligence platform built end to end: eight integrated modules, from risk analysis and network modeling to a partner portal and full API access, on a React front end, a Node and PostgreSQL back end, and a separate Python forecasting service.

ReactNodePostgreSQLPython
Explore the platform

Two ways to work with Peakview Labs

Peakview Insight

Peakview Insight watches your whole network: suppliers, customers, internal sites, and partners. It turns what it finds into prioritized actions rather than a feed of alerts, following best-practice playbooks or your own SOPs so the guidance fits how your operation actually runs. A built-in collaboration portal keeps the response coordinated across your team.

  • Unified view across suppliers, customers, internal sites, and partners
  • AI-powered monitoring with real-time alerts
  • Prioritization and action management, not just detection
  • Tailored recommendations using best practices or your own SOPs
  • Collaboration portal to align responses
Learn more

Peakview Forge

Forge is the method behind the builds above: supply chain software designed from twenty years of operating experience, with AI carrying the implementation so a working system arrives in weeks rather than quarters. Problem framing and architecture stay human. The deliverable is running code you own outright, not a specification.

  • Working software rather than slide decks
  • AI-assisted implementation, human-owned architecture
  • Intelligence built into the system, not bolted alongside it
  • Full code ownership, no vendor lock-in
Get estimate

Technology and Integrations

The stacks behind these builds, and the enterprise systems they are designed to connect to.

AWS
Microsoft Azure
Google Cloud
OpenAI
Anthropic
Google Gemini
SAP
NetSuite
Microsoft Dynamics
Salesforce

Built, Not Theorized

What stands behind the work

Systems Built
0

five prototypes and one full platform, end to end

In Supply Chain
0 yrs

leadership at Amazon, CVS Health, and Clean Harbors

End to End
0%

problem framing through to running code

Want to talk through any of it?

Happy to walk through how any of these systems works, including where the hard parts were.