Business Brain
Teach the business. Review what it learned.
Your price list, return policy and customer exceptions belong in a place you can inspect. Business Brain keeps business facts, rules and procedures with their source, approval state and earlier versions.
A confident reply needs an agreed rule.
“We usually give returning buyers a discount” leaves several questions open: how much, who qualifies, whether bulk orders are different, and when the offer ends. An AI should not fill those gaps by inventing a promise.
Teach POST in the Control Center or Business Brain, then inspect the resulting change. Policies, prices, customer-specific arrangements and other consequential changes wait for an authorized approval. A proposal is not a rule in use.
- Owner-provided knowledge
- Type business information or use supported owner voice input. POST works from the owner’s words, asks about ambiguous details and distinguishes applied information from a pending proposal.
- Documents with evidence
- Upload supported PDFs, Word files, slides, spreadsheets and text. Inspect document processing and extracted facts; retry, re-read or remove a source when needed.
- Rules with a history
- Review proposed changes, edit approved knowledge, stop using an entry and inspect earlier versions. An authorized user can restore a previous version when a policy changes again.
- Customer-aware policy checks
- Supported structured conditions can be evaluated against available customer records. Missing purchase history remains unknown; it should not silently qualify someone for an offer.
Inside the product
Business Brain
Upload a sample wholesale policy, inspect its extracted facts and save an owner-approved business rule.
Real document learning and an owner saving a policy in Business memory. This does not show conversational policy learning.
This recording shows document learning and the business-memory editor. The conversational proposals and rule simulator described here are separate Business Brain controls.
A practical workflow
A distributor sets a returning-buyer discount.
Demonstration example: “Returning buyers receive 10% off. Orders of 80 pieces or more need my review.” The figures illustrate a policy, not POST pricing.
- Give POST the rule with the percentage, customer eligibility and quantity exception stated explicitly.
- Read the proposed changes and approval reasons. Correct any missing or ambiguous condition before approving.
- Approve with an authorized role and use “Try a rule” to inspect returning, unknown-history and bulk-order scenarios.
- When the offer becomes 8%, review the new version and its difference from the previous rule. Keep the earlier version available in history.
Set the controls first.
Business Brain approval and editing are role-controlled. A customer conversation can consult supported policies; it cannot teach the owner’s rules or approve changes.
Check dates, amounts, customer scope and evidence before approval. Voice extraction and multilingual understanding are provider-dependent, so inspect the transcript and proposal.
Check the fit.
- Purchase-history conditions need reliable connected commerce records. Without those records, eligibility can remain unknown and require a person.
- Learned procedures guide a conversation. They do not create executable workflows or authorize an external action.
- Policy evaluation handles supported conditions. Conflicts and unsupported situations can require escalation; checks do not guarantee every kind of customer promise.
Follow the work.
Read the business-policy training guide →Start with a 14-day free trial. Review the plans, watch the product and tell us what your business needs during onboarding.