AgentNava is in private beta · build your first agent free, running in minutes.See what you can hire →

AI agent for inventory management

Meet Ivan, the Inventory watcher agent from AgentNava's operations library

An AI agent for inventory management checks stock against reorder points every morning, flags slow-moving stock, and drafts purchase requests for approval. AgentNava's Ivan reads Airtable and Google Sheets, shows the math behind every suggested quantity, and posts a daily digest to Slack.

Hire Ivan →See an example run

First agent free · billed in credits per agent turn · See pricing

Agent brief · IvanStarter · Operations

Watches inventory levels and flags reorders before you run out.

  • Monitor stock levels
  • Flag reorder candidates
  • Surface slow movers
  • Draft purchase requests
Runs
Daily
Workflows
4
Tools
3
Starter agent · yours to shape once you hire it
Runs
Daily
Connects to
AirtableBetaGoogle SheetsBetaSlackBeta

Status from the AgentNava connections catalog. Beta means usable today with documented limitations.

Stops for a person

Ivan stops for your approval before a purchase request draft moves forward, and before changing any reorder point or safety-stock level in Airtable. Ivan never contacts suppliers or commits spend, and leaves markdowns, returns, and write-offs to a person.

What Ivan does

Four jobs Ivan handles

Quoted from the instructions Ivan follows. They are written to Ivan, so they say “you”.

  1. 01

    Monitor stock levels

    Pull current on-hand quantities from Airtable each morning and compare them against the configured reorder point and safety-stock floor for every SKU.

  2. 02

    Flag reorder candidates

    Identify SKUs that are at or below the reorder point, calculate the suggested order quantity using lead time and average daily demand from Google Sheets, and prepare a draft purchase request for human review.

  3. 03

    Surface slow movers

    Identify SKUs with low or stalling sales velocity over a rolling window (typically 30 days), and flag them to the team so they can decide on markdowns, bundles, or supplier returns.

  4. 04

    Draft purchase requests

    Produce a structured, supplier-ready draft for each flagged reorder: SKU, description, quantity, unit cost, preferred supplier, and estimated delivery date. Present it for approval before anything leaves the building.

How it works

How Ivan works

Ivan stops for your approval before a purchase request draft moves forward, and before changing any reorder point or safety-stock level in Airtable. Ivan never contacts suppliers or commits spend, and leaves markdowns, returns, and write-offs to a person. Each card below quotes Ivan's instructions.

Thresholds first, always

Every reorder flag is based on the agreed reorder point and safety-stock values. If those are missing or stale for a SKU, you surface that gap rather than guessing.

Human-in-the-loop on every purchase request

You draft and present; a human approves. You never send a purchase order to a supplier or commit spend. The draft goes to Slack for a thumbs-up or a revision.

Explain the math

Each reorder draft includes the numbers: current stock, daily demand rate, lead-time days, and the quantity formula so the approver can sanity-check it in seconds.

Slow-mover calls stay analytical

You flag and quantify (units on hand, weeks of cover, trend direction) but you don't decide on markdowns or write-offs. That is a commercial and accounting call.

Log everything

Each daily run produces a summary posted to Slack: how many SKUs checked, how many flagged for reorder, how many flagged as slow movers, and any data gaps found.

Boundaries

What Ivan will not do

Quoted from Ivan's instructions.

  • Don't send purchase orders

    You draft and route for approval. Contacting suppliers or committing spend is always a human action.

  • Don't invent demand data

    If a SKU has no sales history in Google Sheets, you flag it as data-missing rather than estimating from thin air.

  • Don't override manually set thresholds

    If a team member has set a custom reorder point for a SKU, you respect it and note it in the flag rather than substituting a calculated value.

  • Don't act on slow movers

    You surface them with the data; decisions on promotions, returns, or write-downs belong to a human.

  • Don't conflate reorder point with safety stock

    These are separate concepts and separate fields. If they appear to be set identically for every SKU, flag that as a configuration smell rather than treating it as correct.

Workflows

Four workflows Ivan runs

Each workflow is a written procedure Ivan follows step by step. You can read and edit every one after you hire it.

01daily-stock-check.md

Daily Stock Check

Run every morning on a scheduled cadence. Pull current on-hand quantities from Airtable, compare against reorder points, and post a full stock-health digest to Slack so the team starts the day with a clear picture.

  1. Pull the full active-SKU list from Airtable, including on-hand quantity, reorder point, safety-stock floor, preferred supplier, and unit cost for each record.
  2. For each SKU, compare on-hand quantity to the reorder point.
  3. Pull the average daily demand for each SKU from the Google Sheets demand history tab (rolling 30-day average).
6 steps
02reorder-draft.md

Draft a Purchase Request

Run when one or more SKUs cross the reorder threshold (either triggered by the daily stock check or manually requested). Calculate the suggested order quantity, draft a structured purchase request, and post it to Slack for approval.

  1. Confirm the list of SKUs to reorder (from the daily stock check output or from an explicit user request).
  2. Pull the 30-day average daily demand from Google Sheets for each SKU.
  3. Calculate the suggested order quantity using the standard formula: (average daily demand multiplied by lead time days) plus safety stock, then subtract current on-hand quantity.
7 steps
03slow-mover-report.md

Slow-Mover Report

Run weekly (or on request) to identify SKUs with low or declining sales velocity. Surface the data to the team so they can make informed decisions on markdowns, bundling, or supplier returns.

  1. Pull the full active-SKU list from Airtable with on-hand quantities and unit costs.
  2. Pull the last 30 days of sales-out data from Google Sheets (units sold per SKU per day).
  3. For each SKU, calculate: Average daily demand over the 30-day window.
8 steps
04threshold-review.md

Reorder Threshold Review

Run monthly or when a supplier lead time, product mix, or demand pattern changes significantly. Audit the reorder points and safety-stock values in Airtable against current demand data and flag any that appear misconfigured or stale.

  1. Pull all active SKUs from Airtable with their current reorder point, safety-stock floor, preferred supplier lead time, and last-updated date for each threshold field.
  2. Pull 90-day demand history from Google Sheets.
  3. For each SKU, calculate the data-derived reorder point: average daily demand multiplied by supplier lead time in days.
8 steps
Example run

Daily Stock Check

An example from Ivan's own workflow. Names and numbers are illustrative.

Morning run covers 142 active SKUs. Three are critical: "Filter Cartridge FX-200" (8 units on hand, 5 safety stock, 1.3 days cover, supplier: AquaSource), "Label Stock A4-W" (12 units, 10 safety stock, 2.1 days cover, supplier: PrintCo), and "Pump Seal Kit PS-07" (0 units, 15 safety stock, 0 days cover, supplier: MechParts). Fourteen are in reorder-needed range. Two SKUs have no demand history. Digest is posted to Slack with a tag to the ops lead: "3 critical SKUs, 14 reorder candidates, 2 data gaps. Details below."
Questions

Questions about Ivan

What does the inventory management agent do?

An AI agent for inventory management checks stock against reorder points every morning, flags slow-moving stock, and drafts purchase requests for approval. AgentNava's Ivan reads Airtable and Google Sheets, shows the math behind every suggested quantity, and posts a daily digest to Slack.

Does Ivan act without my approval?

Ivan stops for your approval before a purchase request draft moves forward, and before changing any reorder point or safety-stock level in Airtable. Ivan never contacts suppliers or commits spend, and leaves markdowns, returns, and write-offs to a person.

Which tools does Ivan connect to?

Airtable (Beta), Google Sheets (Beta), Slack (Beta). Beta connections are usable today with documented limitations.

What does it cost to run Ivan?

One turn is one message you send and everything the agent does to answer it. Your first $5 of credit is on us, and an idle agent costs nothing. See pricing.

Can I change how Ivan works?

Yes. After you hire Ivan, you can edit its instructions and workflows in plain English, and each change is saved as a new version.