Skip to main content
Most issues with Neoflo operations are identified and resolved by the team before you notice them. This guide covers the scenarios where you might spot something first — what’s likely causing it, what you can check on your side, and when to escalate to Neoflo directly.
For urgent issues affecting live operations, contact your Neoflo account manager directly or email hello@neoflo.ai with URGENT in the subject line.

​
Data and Integration Issues

Likely causes
  • API connection timeout between your source system and Neoflo’s processing queue
  • File delivery failure (e.g., SFTP drop or webhook not triggering)
  • Document format mismatch — the file arrived but couldn’t be parsed
What to check
  1. Confirm that your ERP or source system has actually dispatched the document. Check the outbound queue or delivery log in your system.
  2. Verify that the API endpoint or file transfer destination is reachable and returning a success response.
  3. Check whether the document format matches the agreed specification (file type, naming convention, field structure).
Next stepsIf your source system shows a successful send but the document still isn’t in Neoflo’s queue after 15 minutes, contact the Neoflo integration team with the transaction reference and the timestamp of the attempted delivery.
Likely causes
  • API credentials used to sync results back to your ERP have expired or been rotated
  • A recent data model change in your ERP introduced a field mismatch that blocks the update
What to check
  1. Confirm that the API credentials Neoflo uses to update your ERP are still valid and have not been rotated or revoked.
  2. Check whether any recent changes were made to your ERP’s field structure, required fields, or data configuration that could affect the mapping.
Next stepsIf credentials are valid and no data model changes have occurred, contact the Neoflo integration team with examples of the affected transactions. Include the transaction IDs and the expected destination fields so the team can investigate the sync attempt.
Likely causes
  • Your upstream system is sending the same record more than once — for example, due to a retry loop, a batch job running twice, or a double-posting in the source system.
What to check
  1. Review your source system’s outbound log to see if the same document reference was dispatched multiple times.
  2. Check whether a recent system change, maintenance window, or failed job could have triggered a re-send.
Next stepsNeoflo automatically detects and suppresses exact duplicate submissions in most cases. If duplicates are still appearing in the processing queue after an initial delay, notify the Neoflo team with the affected transaction IDs so they can investigate the duplicate handling for that document type.

​
Processing and Exceptions

Likely causes
  • A vendor has changed their invoice format (new layout, new fields, different line-item structure)
  • New document types are entering the workflow that the model hasn’t been trained on
  • An unusual transaction type — for example, a credit note format or multi-currency invoice — is outside the current model’s confidence threshold
What to checkReview a sample of the exceptions to identify whether they share a common document source, vendor, or format pattern.Next stepsSend the Neoflo team at least five examples of the affected documents along with a note on the expected output. The AI will be updated to handle the new format, and you should see the exception rate return to normal within a few business days.
Likely causes
  • The exception record was submitted without sufficient supporting context for the human team to make a decision
  • Required supporting documents (purchase order, goods receipt, approval email) are missing from the exception record
What to checkReview the exception records that breached SLA and check whether all required supporting data was attached. Common missing items include original POs, delivery confirmations, and approval chain documentation.Next stepsEnsure that all exception submissions going forward include the complete supporting data set defined in your onboarding documentation. For exceptions that have already breached SLA, escalate directly to your Neoflo account manager — they will review whether a billing adjustment applies and can help identify a process fix to prevent recurrence.
Likely causes
  • A new cost center was added to your chart of accounts and the GL coding rules in Neoflo haven’t been updated to reflect it
  • A vendor has been recategorised and the existing mapping no longer applies
  • A one-off transaction type fell outside the current coding rule set
What to checkConfirm whether the GL code mismatch relates to a recently added cost center, a newly onboarded vendor, or a change in your chart of accounts that hasn’t been communicated to Neoflo.Next stepsContact the Neoflo team with the correct GL coding rule and the affected transaction IDs. Configuration changes of this type are applied within one business day. Going forward, notify Neoflo whenever you make changes to your chart of accounts or vendor categories so the rules can be updated proactively.

​
Reporting and Dashboard

Likely causes
  • The dashboard is including transactions that were submitted outside the agreed processing windows (e.g., submitted after the daily cutoff)
  • Some transactions in the period had missing or incomplete data at the point of submission, which paused the SLA clock
What to check
  1. Open the transaction log for the affected period and filter for transactions that breached SLA.
  2. Check the submission timestamps against your agreed processing windows.
  3. Look for any transactions flagged as “pending supporting data” — these may be pausing the SLA clock legitimately.
Next stepsIf the numbers still don’t look right after reviewing the transaction log, contact your Neoflo account team and request a reconciliation. They will walk through each affected transaction with you to confirm whether the SLA breach was attributable to Neoflo, your source system, or a data quality issue.
Likely causes
  • A timing difference — the Neoflo report was generated at a different data cutoff than the ERP view you’re comparing it to
  • Filter mismatch — the Neoflo report is scoped to a different date range, entity, or process type than the ERP report
What to check
  1. Confirm that both reports use the same reporting period start and end dates, and that the data cutoff time aligns.
  2. Check that any entity, cost centre, or process filters applied in the Neoflo dashboard match the filters used in your ERP report.
Next stepsIf the discrepancy persists after aligning the date range and filters, email hello@neoflo.ai with both reports attached and a note on the specific line items that differ. The team will trace the variance to its source and respond within one business day.
Subscribe to the Neoflo weekly operations summary report to catch anomalies before they become issues. The summary highlights exception rate trends, SLA performance, and any data delivery gaps — giving you a regular snapshot of operational health without needing to check the dashboard manually.