Tell the agent what you want — more sales, more profit or more units — and where it must never go. It applies the science underneath: real A/B tests on your shoppers, measured outcomes and a learning loop that compounds with every experiment.
Pick Maximize Sales, Profit or Volume — store-wide, per category or per product — plus margin floors and step limits.
The agent runs rigorous 50/50 price experiments on real shoppers, inside your guardrails only.
Results are judged on cash per visitor with statistical confidence. Each outcome sharpens the agent's model of how your shoppers respond.
Winners roll out, losers revert, and the agent keeps retesting as demand, seasons and competitors change. It learns mathematically and gets better over time.
Reinforcement learning from your own experiments instead of fixed if-then rules.
Margin floors, price limits and charm endings are never crossed. You can override anything, any time.
Every test adds knowledge, so recommendations get more accurate month after month.
Ask questions and give instructions in plain words, inside the app or from Claude and ChatGPT.
Install prAIcing on your smartphone home screen to follow experiments on the go.
Each price change comes with a plain-English reason and one-click rollback.
| Legacy repricer | prAIcing Agent | |
|---|---|---|
| How you instruct it | Write and maintain rules | State a business goal and guardrails |
| Decision basis | Competitor prices | Measured behaviour of your shoppers |
| Proof it works | None — changes are assumed good | Rigorous A/B tests with confidence levels |
| Over time | Same rules, same results | Learns from every test and improves |
| Margin | Erodes in price wars | Protected by your floor |
| Access | Desktop dashboard | Web, phone app, in-app chat, Claude & ChatGPT |
A repricer executes the rules you write. prAIcing receives your goal and guardrails, then decides which experiments to run, measures the results and adapts — continuously, without you scripting each move.
Every price test is feedback. The agent updates its understanding of how demand responds to price for each product and uses it to choose better next tests, so its decisions improve over time.
No. It works strictly inside your guardrails, you choose Observe, Recommend or Autopilot mode, and every change can be rolled back with one click.
Yes. Competitor prices are one input to its tests, not a price to copy blindly.