Jarvis V2

v2.0.0
0/7 Agents
Uptime:
BQ
API
Live
Loading market data...
Shanfeld Equities

Shanfeld Equities AI

“Res Ipsa Loquitur” — The thing speaks for itself

Born from Coastal Oak Capital — a family office built on the belief that the best investment decisions come from consensus, not conviction alone. What started as a personal trading journal evolved into an enterprise-grade AI system where 7 artificial intelligence models debate every decision before a single dollar moves.

The moat isn't the data — it's the architecture. A vertically integrated AI intelligence stack that turns publicly available information into proprietary signal through multi-model consensus and recursive learning.

The Founder

David Shanfeld

Founder & Chief Architect — Shanfeld Equities

David built this platform from scratch over 16 weeks — not by writing every line of code manually, but by architecting the system at a conceptual level and orchestrating AI to execute the vision. Every API integration, every agent configuration, every risk guardrail reflects intentional design decisions about how machines should assist human judgment in financial markets.

His investment philosophy centers on picks and shovels — owning the infrastructure that powers megatrends rather than betting on which company wins the race. Government-backed assets with bipartisan tailwinds. AI infrastructure. Power generation for data centers. Critical minerals. The thesis is simple: follow the money that's already been allocated.

Coastal Oak CapitalEnterprise AIMulti-Agent SystemsPicks & Shovels Thesis

Entity

Shanfeld Equities LLC

Family Office

Coastal Oak Capital

Platform

Jarvis V2

AI Agents

7 Models in Consensus

Data Sources

37+ Live APIs

Build Time

16 Weeks

Build Timeline

1Weeks 1–4

Foundation

Core orchestrator, first 3 AI agents, basic consensus engine, BigQuery integration

2Weeks 5–8

Intelligence

Trinity System, Council of Sages, Devil's Advocate, 31+ data sources connected

3Weeks 9–12

Execution

Strategy engine, tax-loss harvesting, options income, margin management

4Weeks 13–16

Enterprise

Full dashboard, guided tours, decision ledger, accuracy tracking, cloud deployment

System Architecture

Layer 0Data Ingestion

37+ real-time sources: market data, news, social media, EDGAR, congressional trades, options flow, dark pool activity, Substack, Discord

Layer 17-Agent Consensus

Claude Opus, Sonnet, GPT-o1, Gemini, Grok, Perplexity, and Parallel — each analyzes independently, then votes. No single model bias.

Layer 2Trinity System

Three specialized AI primes (QUANT, PSYCH, RISK) vote on opportunities with exponential confidence multipliers. RISK has veto power.

Layer 3Council of Sages

Five legendary investor frameworks (Buffett, Dalio, Marks, Thiel, Munger) validate positions. Howard Marks can veto.

Layer 4Devil's Advocate

Generates the strongest possible counter-thesis for every bullish case. Second-level thinking baked into every decision.

Layer 5Strategy Engine

Automated margin reduction, tax-loss harvesting, options income generation, and allocation rebalancing — all based on your real portfolio.

Trinity System — Second Layer of Validation

After the 7 AI agents reach consensus, three specialized “Prime” agents run a second-level analysis. All three must agree, or the position size shrinks. RISK_PRIME has absolute veto power.

QUANT_PRIME

Quantitative Analysis

Options flow, insider trades, dark pool activity, fundamentals

PSYCH_PRIME

Market Psychology

Market sentiment, behavioral signals, fear/greed cycles

RISK_PRIME

Risk Assessment

Portfolio architecture, position sizing, VETO POWER

VETO POWER

3/3

27x

Full position (5%)

2/3

8x

Half position (3%)

1/3

1x

Watch only

VETO

0x

No position

Council of Sages — Legendary Investor Frameworks

Every trade is evaluated through the lens of five of history's greatest investors. Their frameworks are encoded as AI prompts that challenge each thesis from different angles.

Warren BuffettQuality & Moats
ROE, FCF Yield, Margins
Ray DalioRisk Parity
Correlation, Beta, Diversification
Howard MarksMargin of Safety
P/E, Stress TestsVETO
Peter ThielMonopoly Power
TAM, Market Share, Network Effects
Charlie MungerMental Models
ROIC vs WACC, Accounting Quality

Investment Thesis: Picks & Shovels in the AI Revolution

Don't bet on which gold miner wins — own the shovels, the railway, and the general store. The core thesis focuses on infrastructure plays with government-backed catalysts.

Government-Backed Assets

Highest conviction — CHIPS Act, IRA, DOE funding. Bipartisan money already allocated and flowing.

Conviction: Highest

AI Infrastructure

Semiconductors, data center equipment, networking. The picks and shovels of the AI gold rush.

Conviction: High

Power & Energy

Nuclear renaissance, natural gas, grid infrastructure. AI data centers need massive, reliable power.

Conviction: High

Critical Minerals

Rare earths, lithium, uranium, copper. The TRUE NEEDS that every megatrend depends on.

Conviction: Medium-High

Technology Stack

AI Models

Claude Opus 4.6
Claude Sonnet 4.5
OpenAI o1
Gemini 2.0 Flash
Grok 3
Perplexity Sonar-Pro

Backend

Python / FastAPI
BigQuery
Polygon.io
FRED API
EDGAR / SEC
Plaid Banking

Frontend

Next.js 14
TypeScript
TailwindCSS
React Query
Google Cloud Run
Vercel Deployment

Shanfeld Equities LLC — Enterprise Proprietary AI

Coastal Oak Capital — Family Office

“The moat isn't the data — it's the architecture.”