See the disruption before the email arrives.

Supply chains generate the richest operational data in the enterprise and act on the least of it. We build forecasting, optimization, and document automation that turn that exhaust into earlier decisions — measured in inventory days and service level.

Discuss your use case →What we build

quantpi · industry/telemetry

$ industry.constraints()
✓ metrics: forecast error · inventory days · OTIF · docs/hour
✓ deployment: cloud · hybrid · air-gapped
✓ proof standard: measured on your data
# domain constraints are design inputs, not blockers

IND/01What's at stake

Every planning cycle you run on stale signals is margin you donate.

The gap between when reality changes and when your plan changes is where supply-chain money dies: expedites, stockouts, write-offs. Closing that gap is an engineering problem — demand sensing from live signals, optimization that respects real constraints, and document automation that removes days of clerical latency from goods movement.

IND/02What we build here

Use cases we ship

Demand sensing & forecasting

Forecasts that ingest POS, orders, weather, and market signals — re-planned continuously instead of monthly, measured on MAPE reduction.

cadence: continuous

Inventory optimization

Multi-echelon safety-stock and replenishment optimization against service-level targets — fewer days of inventory at equal or better OTIF.

target: service-level pinned

Trade & logistics document automation

Bills of lading, commercial invoices, customs declarations — extracted, validated, cross-checked against orders, exceptions routed.

throughput: docs/hour

ETA & disruption prediction

Shipment-level arrival prediction with disruption alerts early enough to re-plan — fed by carrier, port, and weather signals.

alerting: pre-disruption

Network & flow analytics

Lane, node, and supplier performance analytics — structural costs and risks made visible, scenario modeling for network changes.

view: network-wide

Supplier risk monitoring

Continuous monitoring of supplier signals — financial, operational, geopolitical — scored against your exposure, not generic indices.

scoring: exposure-weighted

IND/03Domain constraints we design for

Built for your constraints

Integration across ERP, WMS, TMS, and carrier systems
Forecast accuracy measured against naive baselines honestly
Optimization constraints validated by planners before go-live
Document pipelines tolerant of 40+ formats and languages
Alert precision tuned — planner trust is the scarce resource
Cash-cycle impact modeled per initiative

IND/04Questions, answered straight

FAQ

Our data lives in SAP plus a dozen spreadsheets. Realistic starting point?

Completely standard. The first phase builds the harmonization layer — extracting from SAP, normalizing the spreadsheet logic planners actually use, and creating the unified view models need. We've never met a clean supply-chain data estate; the methodology assumes mess.

How much forecast improvement is realistic?

Against monthly statistical baselines, demand sensing typically cuts error 15–30% at SKU-location level — but the honest measure is your baseline, which we benchmark first. Where your current process is already strong, we'll tell you, and point the investment elsewhere.

Can document automation handle customs paperwork across countries?

Yes — multi-language, multi-format processing is the core design requirement. Extraction is validated against order and shipment data, and country-specific rules drive the validation layer. This domain is also where our QEXIM product originates.

How do planners interact with these systems?

Through their existing tools wherever possible — recommendations land in the planning system with reasons attached, and planners accept, adjust, or reject with one action. Override patterns feed back into the models. Adoption is designed, not hoped for.