Automate your DeFi risk process.
Your protocol. Our proof of work.
Submit a protocol. Receive structured, source-backed evidence across your risk framework — in hours, not weeks.
Every field traced to its source. Conflicts flagged with citations. Confidence scored — analysts review exceptions, not every row.
We automate the evidence pipeline, not the judgment. Your risk framework stays yours.
Trace every claim to its source.
Every field cited, every conflict surfaced, every confidence score auditable. Judge the output before committing to anything.
No integration. No meeting. Results in 2 hours.
Not a black box. An evidence chain.
Define which fields matter for your assessment. We run deep, multi-source analysis with cross-verification to research every field in your schema. You get a cited answer to each question you'd ask when evaluating an asset.
Define your schema
Specify the fields you use to make decisions: governance structure, oracle setup, multisig thresholds, or anything else.
Deep research & verification
Multiple AI models use our research engine to search docs, on-chain data, and public sources. Each field is researched independently, then cross-verified across models to catch omissions.
Structured, cited output
Every field gets a researched answer. Your analysts review the flagged conflicts and low-confidence findings, not the entire dataset.
Every finding includes
Exact source quotes
Verbatim excerpts from the original source
Source links
Direct links to docs, contracts, or on-chain data
Conflict detection
Flagged when sources disagree on the same field
Auto-resolution
Conflicts resolved automatically when on-chain data confirms one side
See it in action.
Expand a finding, then click View Sources to see exact quotes, model-by-model verification, and rejected evidence.
Governance docs specify a 48-hour timelock for contract upgrades, but on-chain analysis shows the ProxyAdmin owner can execute upgrades with 0-second delay. This discrepancy means governance safeguards documented for token holders are not enforced at the smart contract level.
On April 15, 2026, 15% of total token supply unlocks for early investors and team members. At current prices, this represents ~$12M in liquid tokens entering the market. Average daily DEX volume is only $1.8M, suggesting potential for significant price impact.
Your rules in. Your proprietary signals out.
Your Risk Framework
Schema Definition
What fields to verify
Criteria Rules
What counts as a risk
Source Priority
Which sources matter
Risk Thresholds
When to flag
Verification Engine
Ingestion
Processing
Synthesis
Verification
Formatting
Integrated data sources
More sources coming
Your Proprietary Risk Signals
Verified Fields
Source-backed data points
Evidence Links
Full citation per field
Conflict Flags
Source vs. source mismatches
API-Ready Output
Structured JSON & sheets
Same engine, different outputs. Your frameworks, criteria & thresholds make results unique. The engine is shared. The IP is yours.
Trace how every answer was built. Every source, every step.
Follow any claim back to its origin: from the research task, through every search and fetched page, down to exact text snippets and saved sources. Click any node to trace the full chain.
Structured output that plugs into your stack.
Versioned, diffable & citation-backed. Every assessment exports as structured JSON & tabular spreadsheets.
API-ready JSON
Every field carries confidence, evidence URLs & conflict flags. Pipe into scoring models or partner APIs.
{
"protocol": "Aave V3",
"assessed": "2025-02-10",
"total_evidence": 147,
"fields": [
{
"name": "tvl_usd",
"value": 12400000000,
"confidence": 0.97,
"evidence": [
{
"url": "https://defillama.com/protocol/aave-v3",
"snippet": "Total Value Locked: $12.4B across..."
},
{
"url": "https://dappradar.com/protocol/aave",
"snippet": "Aave V3 TVL $12.38B, 7d change..."
}
]
},
{
"name": "oracle_provider",
"value": "Chainlink",
"confidence": 0.72,
"conflict": true,
"conflict_note": "Docs vs. governance proposal",
"evidence": [
{
"url": "https://docs.aave.com/oracle-config",
"snippet": "All price feeds sourced via Chain..."
},
{
"url": "https://governance.aave.com/t/pyth-proposal",
"snippet": "Proposal to migrate oracle infra..."
}
]
},
{
"name": "audit_status",
"value": "completed",
"confidence": 0.94,
"evidence": [
{
"url": "https://github.com/aave/audits/certora",
"snippet": "Formal verification complete, no..."
},
{
"url": "https://blog.openzeppelin.com/aave-v3",
"snippet": "All critical findings resolved as..."
}
]
},
"... 244 more fields"
]
}Spreadsheet export
Bring your own spreadsheet format. Our AI agent adapts to your structure, fills every cell with verified, source-backed data & slots into your existing workflows.
The math behind 100x.
Time per protocol
Manual
8–60+ hours depending on complexity
Research.Tech
~2 hours + analyst review
Refresh cadence
Manual
Monthly, at best
Research.Tech
Weekly or on-demand
Coverage
Manual
Sample and skip
Research.Tech
Every asset, every run
Source traceability
Manual
Trust the analyst's memory
Research.Tech
Every field linked to its origin document
Source conflicts
Manual
Buried in analyst notes
Research.Tech
Both sources shown, conflict flagged
Methodology
Manual
Locked in analysts' heads
Research.Tech
Your schema preserved, applied at scale
Output
Manual
Unstructured notes in Notion
Research.Tech
Structured evidence with citations
Your backlog of uncovered assets is growing.
Clear it in 48 hours.
30-minute walkthrough
Bring a protocol or stablecoin. See every claim traced to its source — live.
Skip the meeting
Send your framework & 3 entities. Structured evidence back in 48 hours.
Your protocol. Our proof of work.
Our customers went from assessing 50 protocols to 200+ — same team, zero new hires.