AVM Discovery
VDR analysis + neural anomaly scoring for M&A deal intelligence (5 tools)
How to pay
Pick whichever fits your workflow — you can switch any time.
Subscribe
Free
Predictable monthly cost with included usage. Best for steady, high-volume traffic.
- Unlimited tools within plan limits
- One API key, billed once a month
- Cancel any time
Pay-per-call
$0.01 – $0.02 per call
Charge agents in USDC the moment they call a tool. No subscriptions, no signup — pay only for what you use.
- 5 priced tools available
- Settled in USDC on Base
- No account or API key required
AVM Discovery is the first-stage intelligence layer of the AVM Alpha pipeline. It combines LLM-assisted Virtual Data Room (VDR) analysis with neural anomaly scoring to surface deal-material findings before they reach the quantitative engines.
Modules
- VDR-Insight — Provider-agnostic extraction: builds grounded LLM requests, verifies citations, extracts heuristics
- avm-anomaly (NIM-Morpheus) — Neural anomaly scoring over entity sets; fuses signals into ranked common.Finding candidates
Tools (5)
| Tool | Price | What it does |
|---|---|---|
| vdr_build_extraction_request | $0.01 | Provider-agnostic VDR LLM request |
| vdr_verify_report | $0.02 | Citation grounding + figure traceability |
| vdr_extract_heuristics | $0.01 | Hand hypotheses to LogNexus |
| anomaly_score_entities | $0.02 | Neural anomaly scoring (no $) |
| anomaly_fuse_findings | $0.02 | Neuro-symbolic fusion |
Architecture
Neural (avm-anomaly) and LLM (VDR-Insight) layers detect and hypothesize. The canonical common.Finding is the only object that carries a dollar — set by deterministic code downstream, never a model.
Smoke Check
vdr_extract_heuristics
anomalyscore_entities