The exception workflow
- Upload your exception list — properties where BPO/appraisal diverges from your internal valuation by more than a threshold (typically 15-20%).
- AVMLens returns a multi-source consensus from up to 11 independent AVMs. If most sources agree with the BPO, the BPO is likely right. If they agree with your number, investigate the BPO.
- Geo-risk evidence explains divergence. A property near a newly designated flood zone or Superfund site may legitimately be worth less than the desktop AVM suggests. The geo adjustment quantifies this.
- Confidence scoring tells you how reliable the consensus is. Low confidence (few sources, high variance) means order another BPO. High confidence means you can act on the data.
What gets routed
- BPO > AVM by 20%+: Possible over-valuation. Check if recent nearby sales support the BPO or if comparables are thin.
- AVM > BPO by 20%+: Possible under-valuation. Check if geo-risk factors (flood zone, environmental) explain the discount.
- High AVM variance: Sources disagree significantly. Manual review or field inspection recommended.
- Geo adjustment > 10%: Location risk is material. The risk-adjusted value may better reflect liquidation reality.
Supporting evidence in the output
- Consensus value with low/high range and variance percentage.
- Individual AVM readings (anonymized) so you can see the spread.
- Geo-risk factors with distances, scores, and source attribution.
- PDF report per property for exception files and audit packages.
Who uses this
- Special servicers adjudicating BPO exceptions on non-performing loans.
- Loss mitigation teams evaluating modification or short sale requests against current value.
- REO disposition teams setting list prices with supporting market evidence.
- QC teams auditing appraisal and BPO quality across the portfolio.