Inventory Management: From Stock Aging Reports to Stock Control Within Minutes
Illustrative business scenario · Sample company and dataSee how PersivX turns stock data into days of cover, transfers before purchases, expiry caught before use, and alerts that reach the right person within minutes of a stock event.
Your ERP tells you what is in stock. PersivX tells your team what is about to run out, where else it is, and what has expired, within minutes of it happening.
The Challenge
Altair Polymers compounds engineered plastics in three plants, ships from a finished-goods depot, and holds 2,860 items worth ₹ 312.40Cr, many of them in batches with an expiry date. Before PersivX:
- A stock report a week old. Stock was exported from the ERP every Monday. By Tuesday afternoon it described a store that no longer existed.
- Short at one plant, spare at another. Line 3 at Pune stopped for a catalyst while Hosur held 300 kg of it. Nobody looked, because each plant read its own report.
- Minimums on paper. Minimum stock levels lived in a spreadsheet. Nothing watched them between exports.
- Expiry found too late. Quality checked batch expiry by hand at month end, and found an expired batch already issued to production.
- Old stock with no owner. ₹ 41.80Cr had sat for more than six months. It was in every report and on nobody's list.
Why It Keeps Happening
None of this is a people problem. It comes from three gaps that most stores and planning teams share:
- The ERP records movements, it doesn't watch them. It posts every receipt and issue accurately. Noticing that an issue just took a critical item below its minimum, at one plant, while another has spare, isn't its job.
- Reports run weekly; stock moves by the shift. A weekly or daily report can't catch an event that matters within the hour, so the line finds out first.
- Insight and action live apart. The shortage is in a report, the transfer request in an email, the requisition in the ERP, the approval in someone's inbox. Each hand-off costs hours.
The Cost of Standing Still
- Stopped lines. In this scenario a 1.4-day stock of catalyst met a 6-day supplier lead time: the choice was air freight or downtime, for material another plant already had.
- Working capital in the wrong place. ₹ 41.80Cr, 13.4% of all stock, is over 180 days old, while planners buy in a hurry for the items that are short.
- Expired material used or written off. ₹ 0.94Cr has already expired and ₹ 2.36Cr expires within 30 days. An expired batch reaching production is a quality risk, not just a cost.
- Plant heads decide on last week's stock. By the time the Monday report is read, the store has moved.
Does This Sound Like You?
- Your stock report is an export someone sends round once a week.
- Production has found a shortage before stores did.
- You have bought urgently what another location had in stock.
- Batch expiry is checked by hand, after the fact.
- Nobody owns the stock that hasn't moved in six months.
If two or more are true, the rest of this page is about you.
The Approach
Close each gap with one operating model: see all stock in one place, decide by cover and lead time, act through approval, and let the watching run itself, every few minutes.
All stock, one view
- By age, location, item type and expiry
- Drill from the total to the batch
What runs out first
- Days of cover against lead time
- Spare stock at other locations
Transfer before you buy
- Transfers and requisitions from the row
- Approvals before the ERP changes
Watched every few minutes
- Stock-outs, expired issues and receipts
- Every delivery on record
What Changes
| Before | After PersivX |
|---|---|
| A weekly stock export | One dashboard for every plant and role, reading the ERP's database directly |
| Production finds the shortage first | The planner hears within minutes of an item falling below its minimum |
| Buying what another plant has | Spare stock elsewhere shown on the row, and a transfer requested through an approval |
| Expiry checked at month end | Expiring batches to quality every morning; an expired batch issued is flagged within minutes |
| Old stock with no owner | Each location head gets their own aged-stock list every Monday, with a liquidation route |
| "How long will it last?" | Days of cover at the forecast rate of use, against each supplier's lead time |
Five Capabilities
PersivX's difference isn't more charts. It connects business intelligence, decision-making, workflow execution and communication in one environment, so all five of these work on the same data, in the same place.
Intelligent stock prioritisation
- Days of cover against supplier lead time
- Spare stock at other locations, on the same row
- Expiring batches and slow movers flagged every day
- Unusual consumption flagged for checking
Analytics-to-action workflows
- Request a transfer from the row
- Raise requisitions over selected rows
- Propose slow stock for liquidation
- Each through an approval, then into the ERP
Alerts within minutes
- Below minimum, expired batch issued, short item received, checked every few minutes
- Daily expiry lists and weekly aged-stock lists
- Email and Telegram built in; WhatsApp, Teams, Slack and SMS through connected providers; a record of every delivery
Role-based intelligence
- Each store sees its own locations
- Planners see every plant
- Stock values masked for store staff
AI-assisted insight
- A daily focus area: what aged, what is short, what expires
- Questions in plain language, such as "which items ran short more than twice this quarter?"
What Is AI, and What Isn't
Not every intelligent-looking number is AI. Be precise about which is which: prospects will ask.
| Capability | What it is |
|---|---|
| Below minimum, spare elsewhere, expiring batches, slow movers | Business rules on stock, minimum levels, locations and dates |
| Stock alerts every few minutes | Scheduled checks against the ERP's database: rules, run often |
| Days of cover | Predictive analytics: stock on hand divided by a time-series forecast of daily use |
| Unusual consumption | Machine learning: an anomaly-detection model that compares each item with its own history |
| Daily focus area | Generative AI: a written summary of what changed, from the report's own data |
| Questions in plain language | Conversational AI: answers from the same data |
What Near Real-Time Means Here
Prospects will ask whether this is real time. Be exact:
- Checked every few minutes, not streamed. Each alert is a scheduled check, every 1, 2, 3, 5, 10, 20 or 30 minutes; every five in this scenario. An item that falls below its minimum reaches the planner within about five minutes of the issue being posted, not the instant it happens.
- Read from the ERP's database. For checks this frequent, PersivX reads the ERP's database directly through a read-only connection, so each check sees the stock as it is at that minute. A copy imported into PersivX refreshes at most hourly, which suits history and trends but not these alerts. Your IT team can point the connection at a read replica to keep the load off the ERP.
- Each event is sent once. An alert sends every time its condition holds. So each near-real-time alert reads only what changed since its previous check: a quiet five minutes sends nothing, and an event isn't repeated every five minutes after.
- Only as current as the posting. PersivX sees an issue when it is posted in the ERP. If stores post at the end of the shift, that is when the alert goes.
Days of Cover
Not every item below its minimum is equally urgent. Days of cover divides what is on hand by the forecast daily use, and sets it beside the supplier's lead time. An item whose cover is shorter than its lead time will run out before a new order can arrive: those go to the top. Where another location holds spare stock of the same item, the next step is a transfer, not a purchase.
What Happens Because of the Insights
| The insight | What happens next |
|---|---|
| An issue takes a critical item below its minimum | The planner and buyer are alerted within minutes, with days of cover and any spare stock elsewhere |
| Another plant has spare stock | The planner requests a transfer from the row; once the plant head approves, the transfer order is created in the ERP |
| No plant has spare stock | The planner selects the short items and raises the requisitions in one run, through the purchase head's approval |
| An expired batch is issued to production | Quality and the plant head are told within minutes, usually before the batch is used |
| A short item is received | The planner and stores hear it has arrived, so production can be rescheduled |
| A batch will expire within 30 days | Quality gets it on the morning list, so it is used first |
| Stock passes 180 days without an issue | It lands on the location head's Monday list, with a button to propose it for liquidation |
Follow-Up Playbook
Every row runs on its own. Nothing new means nothing is sent.
| Situation | How often it is checked | Who hears |
|---|---|---|
| Item fell below its minimum | Every 5 minutes | Planner and buyer for the item, with cover and spare stock elsewhere |
| Expired batch issued | Every 5 minutes | Quality head and plant head |
| Short item received | Every 10 minutes | Planner and stores |
| Batches expiring within 30 days | Every morning | Quality, with the list attached |
| Stock over 180 days | Every Monday | Each location head, their own list |
| Every morning | 08:00 | Plant heads and leadership get the board and yesterday's stock report |
The Roadmap: Four Weeks to Running
PersivX reads the ERP's database through a read-only connection. The stock alerts check it every few minutes, on a schedule you set; history for aging trends and forecasts is refreshed nightly. Nothing in the ERP changes except through an approved workflow, which writes back through the ERP's API.
| Week | Outcome |
|---|---|
| 1 | Stock connected; aging, location and expiry views running |
| 2 | Minimum levels and days of cover; a view per location; values masked for store staff |
| 3 | Alerts every few minutes for stock-outs, expired issues and receipts; daily expiry and weekly aged-stock lists; channels connected |
| 4 | Transfer, requisition and liquidation workflows with approval; consumption forecast and unusual-consumption checks; AI focus area |
Business Impact
Fewer stopped lines
- Shortages known within minutes
- Cover measured against lead time
Stock in the right place
- Transfers before purchases
- Aged stock with an owner and a route out
Expiry before use
- Expiring batches used first
- An expired issue flagged within minutes
Approved, and on record
- Every transfer and requisition approved
- Every alert and delivery traceable
Because this scenario is illustrative, it carries no figures. These are the measures a stores and planning team should baseline before go-live and track afterwards. They are what a verified customer case study would report.
| Measure | What moves it |
|---|---|
| Production stoppages for want of material | Alerts within minutes, and cover against lead time |
| Time from a stock-out being posted to the planner knowing | Checks every few minutes instead of a weekly report |
| Urgent purchases of items held at another location | Spare stock elsewhere on the row, and approved transfers |
| Stock over 180 days, and its share of the total | A weekly owner list and a liquidation route |
| Expired stock written off, and expired batches issued | Morning expiry lists and alerts on issue |
| Inventory days or turns | All of the above |
Five Reasons to Choose PersivX
Traditional BI tells you what happened. PersivX helps you understand what is happening, decide what needs attention, and take action, all from one platform.
One environment, not four tools
- Dashboard, decision, approval and alert on the same data
- No email between the shortage and the transfer
Works on the ERP you have
- A read-only connection, no migration
- The ERP changes only through an approved workflow
Your rules, in the open
- Minimums, lead times and checks you can read and change
- AI used where it is AI, and labelled
Control built in
- Approval before stock moves or is bought
- Each store sees its own; every alert recorded
Value in weeks
- The first view in week one
- Alerts every few minutes by week three
How It Compares
| What the team needs | ERP stock report | Dashboard-focused BI | PersivX |
|---|---|---|---|
| See stock by age, location and expiry | Yes | Yes | Yes |
| Know within minutes when an item falls below its minimum | Varies by ERP; often one module, one person | Depends on refresh and alerting setup | Yes, checked every few minutes |
| See spare stock of the same item at other locations | By hand, location by location | With custom modelling | Yes, on the row |
| Request a transfer or requisition from the report, with approval | In the ERP, screen by screen | Usually a separate workflow tool | Yes, from the row |
| Expiring batches to quality, and expired issues flagged | Limited | If someone builds and checks it | Yes, daily and within minutes |
| A record of every alert, per recipient | No | Varies | Yes |
Typical setups. Capabilities vary by product and edition, so compare against the tools the prospect actually uses.
Talking Points
"PersivX doesn't just report your stock. It watches it, and helps you act within minutes."
| They say | You show |
|---|---|
| "Our ERP already has a stock report." | Days of cover, spare stock elsewhere and the five-minute alerts. None of them are a stock report; they are decisions. |
| "We need real time." | What near real-time means here: checked every few minutes against the ERP's database, within about five minutes of a posting. For stock, that is the gap between hearing first and hearing from the line. |
| "Our ERP has reorder alerts." | The same alert, with cover against lead time, spare stock at other plants, an approved transfer from the row, and a record of who was told. |
| "Alerts every five minutes will flood people." | Each check reads only what changed since the previous one. Every event is sent once, and a quiet five minutes sends nothing. |
| "Store staff shouldn't see stock values." | The same report, with values masked for them and each store scoped to its own locations. |
| "Our ERP can't take the load." | Each check is a small read-only query. It can run against a read replica instead of the ERP's main database. |
What Could PersivX Do With Your Stock Data?
Connect your ERP data. See what is about to run out, move stock before you buy it, and catch expiry before it is used.
Technical Details
For evaluators and implementers: how each part is built with PersivX features.
The dashboard, panel by panel
| Panel | Built with |
|---|---|
| Headline tiles (stock value, over 180 days, expiring in 30 days, below minimum) | Card visuals |
| Stock by age group | Aged from the receipt date; a what-if parameter for bucket size, so every visual re-buckets together |
| By expiry category, by location | Bars by expiry date for batch-managed items; locations stacked by age group; custom tooltips carrying item and batch counts, locations and last movement |
| Stock report | Drill-down with a list parameter for the grouping (item then location, location then item, item then batch) |
| Below minimum, with days of cover | A table of items under their minimum; days of cover and spare stock elsewhere as custom attributes; flow buttons on each row |
| Days of cover vs lead time | Stock on hand over a Forecast (Prophet) of daily consumption, beside the supplier lead time |
| Unusual consumption | Outlier Detection (Isolation Forest) on weekly consumption per item |
| Stock events · today | A query-mode table over the stock ledger: the events the alerts send, as they were caught |
| Expiring batches, slow movers | Tables filtered on expiry date, and on age over 180 days with no issue in 90 days |
| Focus area | AI Summarize and Q&A |
Setting up alerts every few minutes
Each near-real-time alert is a scheduler subscription on a table, with Use as alert, the Every Minute frequency and a Minute Granularity of 1, 2, 3, 5, 10, 20 or 30 minutes.
- Use a direct database connection. The report's datasource must connect directly to the ERP's database, read-only, so each check queries current data. An imported datasource refreshes at most hourly.
- Make each alert's table an event window. An alert sends every time its condition holds; there is no memory of what was sent before. So the table behind each alert lists only what happened since the previous check: for example, issues posted in the last five minutes that left the item below its minimum. Match the window to the frequency.
- A failed check is not retried. If one run fails, the events in its window are not re-sent by the next. They still show on the board, and History shows the failure.
- Keep alert tables small. Every check opens the report on the server in a headless browser and evaluates the condition there. A small, fast table keeps that cheap.
- The server must run the scheduler every minute. Minute schedules only fire if the server's scheduler job runs each minute. Confirm this with whoever runs your PersivX server.
| Alert | Setup |
|---|---|
| Fell below minimum | Every Minute · 5 minute · Condition # Items Greater than 0 · recipients from the Planner email and Buyer email columns · Telegram and email |
| Expired batch issued | Every Minute · 5 minute · Condition # Batches Greater than 0 · quality head and plant head |
| Short item received | Every Minute · 10 minute · Condition # Receipts Greater than 0 · recipients from the Planner email and Stores email columns |
When recipients come from a table column, each row's recipients get their own message, about their rows only.
How the daily and weekly follow-ups are set up
| Situation | Setup |
|---|---|
| Expiring within 30 days | Alert on the expiring-batches table · Condition # Batches Greater than 0 · daily · Attach data in excel |
| Aged stock per location | Run individually for each value of a filter = Location · Sender visual mapping location to its head · Attach data in excel · weekly on Monday |
| Morning brief | Whole report · Attach Screenshot? · Attach data in excel with Date to append: Yesterday |
Every delivery, including every alert, is recorded per recipient in the subscription's History.
Channels
Email and Telegram are built in. WhatsApp, Microsoft Teams, Slack, SMS and ERP or ticketing systems connect as webhook channels: GET, or POST with a JSON body whose placeholders are filled from the row. Each provider's API is configured once as a channel.
Actions from the row (Persiv Automate)
Flow buttons run Persiv Automate flows: Request transfer on a row (the plant head approves, then the flow creates the transfer order through the ERP's API), Raise purchase requisitions over the selected rows (the purchase head approves; Run history shows waiting_approval until then, then completed), and Propose liquidation on a slow mover. Rows are re-read on the server as the button's run-as user, so governance applies, and each row's Run history shows what ran and who ran it. Flows run when a person presses the button; the alerts above tell them when to.
Who a subscription runs as, and what happens when it fails
- Runs as its owner. A subscription runs as the person who created it, under that person's row-level rules, whoever receives it. Editing it doesn't change the owner. To hand one over, the new owner recreates it.
- When an owner leaves. The scheduler skips every subscription whose owner has lost access to PersivX or had Scheduler Access turned off. The subscriptions aren't deleted, and resume if access is given back. To keep one running after its owner leaves, someone else has to create it again.
- Failures are recorded, not retried. A run that fails is marked failed in the subscription monitor, and History shows each recipient that failed and why. There is no automatic retry: fix the cause and use Run now, or wait for the next scheduled run.
Security
Row-level security scopes each store to its own locations; planners and plant heads see every plant. Stock values and costs can be masked for store staff with column masking. Both are applied in the query. The database connection is read-only; the ERP changes only through approved flows that call its API.
Features used
- Direct database connections and scheduled refresh
- Visuals, tables and drill-down
- Custom attributes
- Forecast and outlier models
- Scheduler and smart alerts, including minute schedules
- Flow buttons (Persiv Automate)
- Row-level security and column masking
- AI summaries and Q&A
