Your data, your playbook, one agent, action in the store.
SERA replaces nothing. It connects to what you have, learns how you want stores run, and talks to the people on the floor. Here is exactly what goes in, what happens every day, and what comes out.
Your data in
POS: daily revenue, transactions, units, footfall per store. ERP: stock by SKU per store, deliveries, prices. HR: who works where, start and leave dates. Through an API, a scheduled export or a manual upload at the start.
Your playbook
Policies, SOPs, brand standards, visual merchandising guides, training decks and the commercial calendar are uploaded once. SERA indexes them and uses them as the only source for procedure questions.
Daily analysis
Every morning SERA compares each store with last year, target and the network on footfall, conversion, revenue, basket, units per transaction, weeks of cover per SKU and campaign execution, and ranks what needs attention.
Action in the store
A proactive message to the store manager with the cause and one action, a task with a deadline and photo proof, a coaching module, or a reply to any question the team asks in chat. Then a follow-up the next day.
The KPIs, defined the way your finance team would define them.
Ratios are always recomputed from the raw numbers for the period you look at, never averaged from daily percentages, so a weekly conversion rate is the week's transactions divided by the week's footfall.
| KPI | How it is calculated | Compared with |
|---|---|---|
| Footfall | Visitors counted entering the store | Last year, target, network |
| Conversion | Transactions ÷ footfall | Last year, target, network, store rank |
| Revenue, average basket, units per transaction | From the receipt lines; basket = revenue ÷ transactions, UPT = units ÷ transactions | Last year (same date), target (same weekday last year × growth) |
| Weeks of cover | Closing stock ÷ average weekly sales of the last four weeks, per SKU and store | Low-stock list (worst ten), overstock list by value, dead stock |
| Campaign execution | Sales lift on campaign SKUs vs the previous period, photo proof of the display | Stores that executed vs stores that did not |
| Top sellers and top products | Receipt lines per seller, units per product | Store and network |
Plays with your stack.
SERA reads data; it does not write back into your POS or ERP. That keeps the integration small and the risk low.
- REST API or a scheduled CSV/Excel export, daily at store level and SKU level
- Works with any POS or ERP that can export sales by day and stock by SKU; we have templates for the common ones
- HR data limited to name, store, role, start and leave date
- Historical data for the last 12 to 24 months loads once so "versus last year" works from day one
- Push notifications to the mobile app; WhatsApp and Teams delivery on the roadmap
Security and hosting
EU hosting by default. Local hosting on request for regional deployments. The client can run SERA on its own OpenAI API account so conversation data never leaves an account it controls.
Each partner (brand, franchisee, subsidiary) has its own data space, its own prompt and tone, and its own admin users. Store users see only their store; area managers see their area.
Every conversation and alert is logged in the admin portal. Token usage is tracked per partner.
From signature to the first store in two to three weeks.
Demo on your KPIs
45 minutes with your own definitions, your store list and one month of data if you have it.
Data and playbook
Connect the export, load last year's history, upload SOPs and standards, set the tone of voice.
Pilot stores live
Store managers onboarded in a 20-minute session. Proactive alerts switched on. Weekly review with you.
Decision
You see conversion, task completion and usage per store before you pay. Then roll out or stop.