● Live — v0.1.0

Financial intelligence
that actually works.

The AI analyst desk for stocks, options, and crypto27+ external services, 26 specialists, one deterministic engine. MarketIntell doesn't hallucinate — it computes.

27+
External Services
26
Specialists
1
LLM Call / Query
23 Wins. 2 Losses. 1 Tie.

Blind evaluation by Claude on 26 real financial queries. Category scores rated 1–10.

Polymarket bot · 11/11 markets won · +$981 real USDC · on-chain
Technical Analysis
9.0vs4.0
Market State
9.0vs1.0
Trenches / Degen
8.5vs1.0
Microstructure
8.0vs3.5
Fundamentals
8.0vs1.0
Macro / Regime
8.0vs3.5
Polymarket
8.0vs1.0
JSON Output
8.0vs5.0
Social / Sentiment
7.5vs4.5
Multi-Domain
8.6vs2.6
8.8
Avg Data Quality
8.3
Avg Synthesis
7.8
Avg Actionability
Built for traders, not chatbots.

Every component is purpose-built for financial intelligence — from data ingestion to reasoning to delivery.

🧠

Chain-of-Thought Engine

Single LLM call for synthesis. Everything else — data fetching, classification, routing — is deterministic. No agent loops. No hallucinated tool calls.

📡

27+ External Services

CoinGecko, Coinglass, DeFiLlama, Nansen, LunarCrush, Polymarket, FinancialDatasets, TwelveData, Massive, Yahoo Finance, Twitter/X — all fetched in real time with circuit breakers and caching.

🎯

26 Specialists

Technical, Microstructure, Fundamentals, Macro, Social, Polymarket, Smart Money, Options Flow, Earnings, Sector — each with dedicated data pipelines and reasoning logic.

🌍

Stocks, Options, Crypto

Stocks, options, and crypto in one API. Cross-asset correlation. Ask about BTC and NVIDIA in the same query.

Alpha Signals

Autonomous signal generation with accuracy tracking. Backtesting engine. Screeners for momentum and fundamentals. Webhook delivery.

🤖

AI Agents

Agents with persistent memory, scheduled execution, and approval workflows. Define your strategy, let it run autonomously.

🧩

Structured Output

Get responses as typed JSON via response_format. Built-in schemas for trade_setup, market_analysis, and token_screen — or pass your own JSON schema.

Technical Analysis Microstructure Fundamentals Macro / Regime Social Sentiment Polymarket Smart Money Crypto Stocks Options
One query. Full analysis.

Ask a question in natural language. Get real-time data, multi-domain synthesis, and actionable intelligence.

  Request — POST /v1/chat
// Natural language query { "message": "What's the technical setup
on ETH right now? Include
funding rates and any smart
money moves."
, "stream": false }
  Response — 3 specialists, 12.4s
{ "specialists_used": [ "technical", "microstructure", "smart_money" ], "content": "## ETH Technical Setup
Price: $3,847 | 24h: +2.4%
RSI(14): 62.3 — Neutral-Bull
MACD: Bullish crossover (2h)

## Funding & OI
Perp funding: +0.012% (mild long)
OI delta 24h: +$340M

## Smart Money
3 wallets accumulated 12.4K ETH
in last 6h via DEX aggregators
..."
, "latency_ms": 12400 }
How a query flows.

One LLM call. Everything else is deterministic code — no agent chains, no retry loops, no hallucinated tool use.

01

Classify

Deterministic parser identifies tickers, intent, and required specialist domains

02

Fetch

Parallel data retrieval from 27+ external services with circuit breakers and caching

03

Compute

Specialist engines run technical indicators, funding analysis, on-chain metrics

04

Synthesize

Single LLM call merges all specialist outputs into a coherent, actionable response

Use it how you want.

REST API, Telegram bot, Claude Desktop via MCP, or the built-in web chat. Same engine everywhere.

⚙️

REST API

OpenAPI spec. SSE streaming. Webhook delivery. Full programmatic access.

💬

Telegram Bot

@MarketIntellStagingBot — ask questions, get signals, manage agents from chat.

🔮

MCP (Claude Desktop)

Native integration with Claude Desktop. Use MarketIntell as a tool inside Claude.

🖥️

Web UI

Built-in chat interface with session history, streaming, and specialist breakdowns.

Stop asking ChatGPT
about markets.

General-purpose AI scores 2.6/10 on financial queries.
MarketIntell scores 8.6. The data speaks.