Every order ships from one accurate picture of inventory
We build the software layer that keeps a distributor honest about what it has and what left the building. The problem is rarely the warehouse — inventory lives in an ERP, a webstore, a marketplace, and a picker's memory, and those numbers stop agreeing by Tuesday.
We start at the transaction, not the dashboard: an order traced from channel to ERP to pick list to carrier label to invoice, with every retyping point closed. Event-driven inventory sync with idempotent writes, scanner confirmation at pick and pack, EDI mapped once and monitored after.
A first project is usually small and close to the money: the order path from one channel into the ERP with no retyping, or a receiving flow that puts a scan between the dock and the on-hand number. Distribution margin does not leave room for a long discovery phase, so we pick the leak that is costing chargebacks or oversells right now, close it, and let the next project follow from what the first one exposes. The integration layer built for the first channel is the same one the second and third reuse.
Reviewed
Sound familiar?
If you've said any of these, we should talk.
“We sold three of something we had one of.”
Oversells come from stale inventory pushes, not careless people. We publish availability the moment a pick, receipt, or adjustment commits, hold reservations against open orders, and run a reconciliation job that reports drift.
“Onboarding a new EDI trading partner takes us months.”
Most of that is hand-mapping and testing blind. We build a partner-agnostic document model with per-partner maps, a harness replaying real 850s and 856s before go-live, and alerting so a rejected ASN surfaces in minutes.
“Our pickers walk the whole building for one order.”
Slotting is a data problem before a shelving problem. We analyze pick history by velocity, affinity, and cube, propose slot moves with the travel time each saves, and build pick-path logic into the scanner flow.
“Cycle counts take a weekend and still do not match.”
We build scanner-based cycle counting into the working day: tasks issued by ABC class, variance captured with reason and person, a running accuracy metric by zone. Exceptions point at the process that caused them.
“Backorders live in a spreadsheet, and the customer finds out when the box arrives short.”
Backorder handling belongs in the order flow, not beside it. We hold the open line against inbound receipts, allocate automatically when stock lands, and notify the customer with a real date when it changes. Sales sees the same allocation the warehouse does, so nobody promises stock twice.
How this industry actually runs
The operation as we usually find it.
A distributor runs on systems never designed to talk. An ERP or WMS holds inventory and orders; a webstore and marketplaces sell against the same stock; trading partners send 850 purchase orders and expect 856 ASNs and 810 invoices back in their own flavor of the standard. The floor day is receiving and putaway, waves of picks, pack and manifest against carrier rates, and returns. Slotting decides how far a picker walks. Lot and serial capture decides whether a recall is an afternoon or a month. Margin is thin enough that a chargeback over a late ASN is the quarter.
Assets, orders, and people on one picture
Components:
- Assets & fleet (telematics, meters)
- Orders & tickets (ERP, portals, email)
- Integration hub (documented interfaces)
- Ops dashboard (exceptions first)
- Field app (dispatch, proof of work)
Connections:
- Assets & fleet to Integration hub (telemetry)
- Orders & tickets to Integration hub
- Integration hub to Ops dashboard
- Integration hub to Field app
What we build
Starting projects that fit Distribution & Wholesale.
- Inventory sync across ERP, webstore, and marketplaces with drift reporting
- EDI for 850, 855, 856, 810, and 940/945 with failure alerting
- Scanner workflows for receiving, putaway, pick, pack, and cycle count
- Order orchestration with carrier rate shopping, labels, and tracking writeback
- Slotting and pick-path analysis built from your own pick history
- Lot, serial, and expiry capture with audit-ready traceability
- Customer portals with real availability, pricing tiers, and reorder history
- Purchase order intake that reads emailed and PDF orders into ERP-ready records, with human approval before the write
Capabilities we bring
What we'd build first
Three places a first project usually starts.
01
Inventory sync across ERP, webstore, and marketplaces with drift reporting
02
EDI for 850, 855, 856, 810, and 940/945 with failure alerting
03
Scanner workflows for receiving, putaway, pick, pack, and cycle count
Scoped in writing before any work starts — how engagements work.
Working in Distribution & Wholesale?
Tell us the line.
What runs by hand, what is not connected, what you are trying to build. An engineer replies within one business day with whether and how we would approach it.
Common questions
What Distribution & Wholesale teams ask first.
Do we have to replace our ERP or WMS?
Almost never. Most distribution problems are integration and workflow problems on top of a system that holds the data fine. We would rather add a scanner flow and a sync service than sell a replacement.
Can you work with our marketplace feeds and 3PL?
Yes. Marketplace APIs, 3PL flat files, and partner EDI all end up in one document pipeline with validation, retries, and visible failures. The formats differ; the engineering underneath does not.
Where does AI actually help a distributor?
Where a person reads unstructured text all day: parsing PDF purchase orders into order records, matching supplier part numbers to yours. Anything writing to the ERP sits behind human approval until the error rate earns autonomy.
How long does an inventory sync or EDI project usually take?
It depends on how many systems and partners are in scope, and how clean the data is. A single-channel sync or one trading partner is typically measured in weeks rather than months, most of it spent on edge cases: partial shipments, substitutions, units of measure. We scope the first slice narrowly, get it live with monitoring, then add channels and partners against the same pipeline. We quote a range after seeing the systems, not before.
What happens when the sync breaks at 2 a.m.?
It will, eventually — a partner changes a field, an API rate-limits, a marketplace returns an error nobody has seen. We design for that: every document and inventory event is logged with its state, failures retry with backoff, and anything that exhausts retries lands in an exception queue with an alert to a named person. Usually the fix is a replay, not a rebuild. We typically hand over with runbooks and offer ongoing support if you want us on call.
More in Operations & Field
Strategy. Software. Systems.
Engineering for Distribution & Wholesale.
Describe the problem in your own words. An engineer reads it — not a sales script — and tells you plainly what it would take.