Priority: Medium
Feature description:
The existing dispute module requires reps to fill a structured form. Enhance it with a freeform text input that uses an LLM to extract the disputed transaction, infer the likely discrepancy type, and pre-fill the structured fields — then route the case to the right approver automatically.
Proposed behavior:
- Rep types: "I think my Q2 deal with Acme Corp was credited to the wrong territory."
- AI extracts: transaction reference, dispute type (crediting), suggested adjustment.
- Pre-populates the dispute form with extracted values; rep confirms before submitting.
- AI also checks if a similar dispute was filed and resolved before, and surfaces the resolution as a self-serve suggestion.
- Reduces average dispute submission time and improves routing accuracy.
Why it matters:
Poorly structured disputes create back-and-forth between reps and comp admins. AI pre-triage reduces resolution time and admin burden while making the experience feel modern.
Priority: Medium
Feature description:
The existing dispute module requires reps to fill a structured form. Enhance it with a freeform text input that uses an LLM to extract the disputed transaction, infer the likely discrepancy type, and pre-fill the structured fields — then route the case to the right approver automatically.
Proposed behavior:
Why it matters:
Poorly structured disputes create back-and-forth between reps and comp admins. AI pre-triage reduces resolution time and admin burden while making the experience feel modern.