Industry
Map industry structure, cycle position, and competitive forces
Built for evidence-led investing
Research,
rooted in evidence
Redwood is a local-first U.S. equity research platform. It turns scattered filings, market data, and private knowledge into conclusions you can trace and verify.
Local-first Evidence-auditable No order execution
COMPANY UNDER REVIEW
NVIDIA / NVDA
AS OF
2026.08.31
01
Fundamentals
Statements · Quality · ROIC
02
Valuation
DCF · Scenarios · Sensitivity
03
Market Pulse
News · Technicals · Volatility
EVIDENCE CHAIN
10-K / Revenuematched
Earnings call / Guidancematched
Local thesis / Moatreview
RESEARCH DEPTH
5+2
Core modules
+ two-sided debate
04Local knowledge domains
05+2Company research framework
100%Traceable evidence
0Trade execution
01 / RESEARCH SYSTEMNot more information.
Not more information.
A better research order.
Redwood examines a company through a 5+2 framework: five objective research modules, followed by a two-sided stress test of the case for—and against—investing.
Business model
Deconstruct revenue engines, moats, and unit economics
Management
Assess capital allocation, incentives, and execution history
Financials
Interrogate growth quality, cash flow, and the balance sheet
Valuation
Build price discipline through DCF and relative valuation
+Investment logic
What must remain true? What is the market missing?
−Reasons not to invest
What would invalidate the thesis? Which risks are underpriced?
02 / PLATFORM CAPABILITIESEvery conclusion
Every conclusion
has a provenance.
From the moment information enters Redwood, its source, version, rights, and location are recorded. Agents find connections; humans retain judgment.
01
Local knowledge engine
Turn PDFs, Markdown, DOCX, and private notes into precisely located evidence while preserving source, time, and usage rights.
02
Specialist research agents
Fundamentals, valuation, news, technicals, and volatility work in their own lanes while a coordinator enforces one evidence standard.
03
Deterministic finance
Financial ratios, option Greeks, implied volatility, and risk metrics run through reproducible calculations with inspectable outputs.
04
Financial research ontology
Unify companies, securities, metrics, and relationships while keeping matched, ambiguous, and unresolved identities explicit.
05
Auditable evidence chain
Every material claim can resolve to an original document, precise locator, and evidence class—ready to challenge, verify, and update.
06
Read-only safety boundary
Redwood supports research and risk understanding, never order execution. Sensitive credentials stay outside model context and outputs.
AI can accelerate research.
It cannot replace accountability.
01
Evidence before opinionClaims begin with sources, not confidence.
02
Uncertainty stays visibleUnknowns are never packaged as certainty.
03
Local by defaultPrivate knowledge and sensitive data remain local.
04
Human decidesThe system supports research; the investor decides.
PRIVATE BETA / 2026Built for investors
Built for investors
who go deep.
Redwood is evolving in private beta with investors who value evidence, process, and long-term thinking.