Suggested match
Explanation
Generated from the read-only evidence below. It never changes a match or a record.
The bank narration contains a UTR that is a near‑match (two characters off) to the settlement UTR, likely a typographical error.
The engine identified the transaction amount ₹56,008.51 and date 2026-10-12 as exact matches, but the UTR in the narration (AXISN26252923034) differs by an edit distance of 2 from the settlement UTR (AXISN26285923034). This fuzzy match triggered a suggested match exception with a confidence of 0.875. A human should verify whether the discrepancy is indeed a typo before confirming the match.
Review the UTR discrepancy; if it is confirmed as a typo, approve the suggested match, otherwise reject it.
Evidence records
Read-only source records involved in this exception.
- txn date
- 2026-10-12
- narration
- NEFT CR-AXISN26252923034-GATEWAY SETTLEMENTS PVT LTD-SETTL MIDG3NF3BFV2PY
- credit
- ₹56,008.51
- debit
- ₹0.00
- balance
- ₹1,25,02,769.36
- utr
- AXISN26285923034
- amount
- ₹56,008.51
- settled at
- 2026-10-12
- created at
- 2026-10-12T09:13:00+05:30
- item count
- 17
- fee total
- ₹1,000.84
- gst total
- ₹180.13
Rule trace
The deterministic rule that fired and what it saw.
{
"rule": "stage2.fuzzy_suggest",
"reason": "edit_distance_2",
"best": {
"settlement_id": "setl_zFZETndJefHGkq",
"utr": "AXISN26285923034",
"token": "AXISN26252923034",
"token_source": "narration",
"distance": 2,
"truncated": false,
"confidence": 0.875,
"amount_equal": true
},
"runner_up": null,
"margin": 0.875,
"suggested": {
"settlement_id": "setl_zFZETndJefHGkq",
"utr": "AXISN26285923034",
"token": "AXISN26252923034",
"token_source": "narration",
"distance": 2,
"truncated": false,
"confidence": 0.875,
"amount_equal": true
},
"candidates": [
{
"settlement_id": "setl_zFZETndJefHGkq",
"utr": "AXISN26285923034",
"token": "AXISN26252923034",
"token_source": "narration",
"distance": 2,
"truncated": false,
"confidence": 0.875,
"amount_equal": true
}
]
}Audit history
Append-only. Actor: system, llm or human.
- llminvestigation
source=llm · model=openai/gpt-oss-120b · review_policy_applied=false
- systemjournal draft
applicable=false · reason=SUGGESTED_MATCH needs a matching decision, not a journal entry