Excess Inventory01

Excess inventory, triaged and disposed of by rule.

We encode your disposal rules — last-sale and last-receipt evidence, vendor return windows, markdown-vs-return — inside the POS/ERP your stores already run.

  • AI implementation services
  • Your disposal rules, encoded
  • Exception-only queue for buyers
  • No new system for your stores

Official services partner of the platforms defining AI

NVIDIAAnthropic
Peter Enestrom, founder of Zaigo

Every engagement is led personally by Peter Enestrom and the Zaigo AI & engineering team

YaleColumbia UniversityMicrosoft
The untouched pile01

Everyone can see the pile. Nobody owns the decision.

From hardware and building-supply retailers we have worked inside: 19–22-store operations on Epicor Eagle-class POS/ERP, where dead and discontinued stock sits untouched because no one triages it.

  • Product no one will buy, store after store

    “Stores filled with hundreds of thousands of dollars of product no one will buy.” Dead and discontinued inventory sits untouched because no one triages it.

  • Special orders never picked up

    “Special-order items never picked up, invisible and unsellable.” End-of-life and special-order stock disappears in the POS, and the 30-day vendor return window closes before anyone flags it.

  • Markdowns nobody verified

    “$740K of markdown inventory stuck at 90% off and nobody verified the tags ever got hung.” Markdown approved; markdown compliance, the price reaching the shelf, never checked.

  • Discontinuing blind

    “I’m committing to leaving a shelf space empty with no work done to understand the impact.” Items discontinued with no last-receipt or last-sale data attached.

  • Purge alerts without an audit trail

    “Dead and slow-moving SKUs for purge alerts.” Operators want dead-inventory flags for store hygiene — and every system change logged for audit, not a spreadsheet someone edited.

What manual costs02

The pile compounds while the window closes.

Measured inside multi-store hardware retail operations. Your numbers will differ — the triage audit puts figures on one store and one category before anything gets built.

$740KMarkdown inventory stuck at 90% off, tags never verified hung
30 daysVendor return and credit window, missed while the pile sits
30–40%Count error rates on markdown and older items

From anonymized engagements — 19–22-store hardware chains, Epicor Eagle-class POS

How it works03

We encode your disposal rules.

No new system for your buyers to learn. The rules your best merchant carries in their head become the triage every flagged SKU goes through.

  1. 01

    Triage audit

    One store, one category. We quantify the excess — last-sale and last-receipt evidence per SKU, special orders never picked up — before anything changes.

    Weeks 1–2
  2. 02

    Encode the disposal rules

    Markdown-vs-return per SKU, vendor return windows with days-left flags, special-order sweeps, purge alerts with an audit trail — encoded inside the POS/ERP you already run.

    Weeks 3–8
  3. 03

    Exception-only queue

    The rules mark down, return, or write off with the reason logged. Buyers see only what the rules can’t dispose of — a short queue, not the pile.

    Ongoing
Not a liquidation service

A liquidation marketplace moves the pile once. It can’t stop the next one.

The software sells another report of the pile: the same dead SKUs every Monday, none disposed of. Tools ship with generic rules; your stores run on exceptions — $50 markdown thresholds, 30-day vendor return windows.

We are neither. We encode your disposal rules into the systems your stores already run — our AI reads every flagged SKU against those rules — so the pile drains on a schedule and the next one never forms.

In production
30 daysVendor return windows now flagged with days left, not missed
$50Markdown approval thresholds enforced by rule, not by memory
1 storeOne store, one category — the triage audit that scopes the build
Dead and slow-moving SKUs for purge alerts — that’s what we asked for. Now the flags come with the last-sale evidence attached, and every change the system makes is logged.
Operations executive, regional hardware store group
Peter Enestrom, founder of Zaigo
Who builds it

Led by Peter Enestrom.

Founder — leads AI & Engineering

Pete Enestrom

Every engagement is led personally by Pete, working with the Zaigo AI & engineering team from the two-week audit through the production handover. The person who scopes the work is the person who builds it.

Education
Yale & ColumbiaGraduate
Background
Microsoft & IntelFormer
Experience
Exited FounderVenture-Backed

Background

Questions04

Asked by owners and merchants.

The straight answers, before you book anything.

Excess inventory is stock on hand beyond what demand will absorb in a reasonable selling period. In practice it is the pile of dead, discontinued, and overbought product that no one has triaged — real inventory records with no next action attached.

Excess inventory can still sell — there is demand, just less than the supply on hand. Obsolete inventory has no demand left: discontinued items, end-of-life SKUs, special orders never picked up. Excess stock gets marked down or returned; obsolete stock gets returned, liquidated, or written off.

With evidence, not feel. Each flagged SKU carries its last-sale date, last-receipt date, and months of supply on hand. Items with no sale since a threshold, no receipt movement, or a bin that cannot be located go to the triage queue — every flag logged so the list survives an audit.

It is a per-SKU decision with a clock on it. If the vendor return window is still open, returning at cost usually beats marking down to 90% off. Once the window closes, the markdown path applies — with the approval thresholds and tag-verification checks encoded so the price actually reaches the shelf.

When the return window has closed, the markdown floor has been reached, and no liquidation channel will take it. The write-off should be the last stop of an encoded rule chain — with the reason, the evidence, and the approval logged — not a quiet adjustment at year-end.

Start with one workflow.

Tell us where your team loses hours. We will come back with a straight answer on whether AI can help, what it would take, and what it would pay.