nasaq.id

Back to case studies
Live production2026Dibangun dalam 3 minggu untuk SoSoValue Buildathon 2026

SignalFlow Agent — AI Trading Signal Dashboard

AI trading signal dashboard — 5-Layer Signal Engine V2, multi-timeframe confluence, paper futures trading. SoSoValue Buildathon 2026 submission.

Context

A custom build where the business flow mattered more than plugging in a generic template.

Next.jsTypeScriptPrismaPostgreSQLSoSoValueTradingAIPWA

Project at a glance

SignalFlow Agent mengubah multi-dimensional market data menjadi trade signals yang explainable. Problem: trader retail butuh sistem yang bisa analisis multi-factor (trend, momentum, volatility, volume, structure) secara otomatis dan execute paper trades — bukan cuma chart viewer. Solution: 5-Layer Signal Engine V2 dengan market regime detection, 7-tier classification, dan wallet-aware paper futures validation. Bloomberg-style command center dengan pipeline visualization.

What this case helps prove

SoSoValue Buildathon 2026 submission by NoHype Labs
End-to-end signal pipeline: market data → multi-factor analysis → classified signal → paper execution
SignalFlow Agent — AI Trading Signal Dashboard screenshot

The challenge

Trader retail kesulitan mengintegrasikan data dari banyak sumber (ETF flows, sentiment, macro events, technical indicators) menjadi satu keputusan trading yang koheren. Platform existing hanya chart viewer tanpa signal classification atau automated execution.

Data tersebar di banyak platform — tidak ada unified view
Tidak ada sistem multi-factor confluence yang explainable
Paper trading terpisah dari analysis — tidak ada feedback loop
Tidak ada adaptasi per trading style (scalper vs swing vs position)

The response

5-Layer Signal Engine V2 yang menggabungkan Trend, Momentum, Volatility, Volume, dan Structure analysis dengan market regime detection. Setiap signal dianalisis di 3 timeframe (1H, 4H, 1D) dengan alignment scoring. Paper futures trading dengan virtual USDC, auto TP/SL/liquidation, dan per-type performance stats.

Why it mattered

SoSoValue Buildathon 2026 submission by NoHype Labs

Before and after, in clearer terms

The point of this section is not decoration. It is to show the operational shift in terms the client team can immediately recognize.

Signal analysis

Before

Manual cek 5+ indikator di berbagai platform

After

Auto 5-factor confluence + 3-timeframe alignment

Trade execution

Before

Manual open/close, rawan emosi

After

Paper futures dengan auto TP/SL/liquidation

Performance tracking

Before

Tidak ada atau manual spreadsheet

After

Per-type stats: Sharpe, Sortino, Calmar, drawdown

Key highlights of the build

5-Layer Signal Engine V2

Multi-factor confluence system: Trend, Momentum, Volatility, Volume, Structure — dengan market regime detection, 7-tier classification, dan volatility-adjusted TP/SL. Adaptif ke 4 trading styles (Scalper, Intraday, Swing, Position)

Multi-Timeframe Confluence

Setiap signal dianalisis independent di 3 timeframe (1H, 4H, 1D) dengan alignment scoring. Semua 3 timeframe agree = 95 conviction score, conflicting = 30-50

End-to-End Signal-to-Execution

Bloomberg-style command center dengan pipeline visualization. Paper futures trading dengan virtual USDC, auto TP/SL/liquidation, leverage 1x-100x, dan per-type performance stats

Impact that stayed visible

SoSoValue Buildathon 2026 submission by NoHype Labs
End-to-end signal pipeline: market data → multi-factor analysis → classified signal → paper execution

What the system includes

5-factor confluence engine: Trend (EMA/ADX), Momentum (RSI/MACD/ROC), Volatility (BB/ATR), Volume (OBV), Structure (S/R + Fibonacci)
Market regime detection: Trending Up/Down, Ranging, Volatile, Breakout
7-tier signal classification: Strong Long to Strong Short
Trading type adaptation dengan per-type weights, TP/SL multipliers, confidence thresholds
Paper futures trading dengan virtual USDC dan leverage 1x-100x
Auto TP/SL/Liquidation close pada price ticks
Per-type performance stats dan analytics (Sharpe, Sortino, Calmar, drawdown)
MetaMask + WalletConnect v2 pada ValueChain Mainnet
AI signal enrichment dari DeepSeek, OpenAI, OpenRouter (privacy-first, local API keys)
SoSoValue API integration: ETF flows, sentiment, macro events, BTC treasuries
SoDEX Spot + Perps API: live tickers, klines, orderbook, funding rate, open interest
Walk-forward backtest engine dengan per-regime accuracy breakdown
Command-center dashboard dengan pipeline rail, Decision Score, Catalyst Monitor
Alerting system dengan browser notifications
Trade journaling dengan mood tracking
Onboarding modal untuk trading style selection
PWA support dengan custom icons dan Apple Web App metadata
Responsive navigation dengan bottom tabs dan mobile drawer

Technology stack

Frontend

Next.js 16React 19TypeScript 5Tailwind CSS v4Lightweight ChartsTanStack QueryFramer MotionReact Three Fiber

Backend & Database

Next.js API RoutesPrisma ORMPostgreSQL

Integrations

SoSoValue APISoDEX APIDeepSeek / OpenAI / OpenRouterWagmi v3 + ViemWalletConnect v2MetaMask

Architecture

5-Layer Signal Engine V2Walk-forward Backtest EnginePWAZodVercel

If this kind of operational shift is what you need, start from the problem.

The fastest useful next step is not asking for a random feature list. It is clarifying what feels weak today, what friction the team still carries, and what a cleaner first version should solve.

nasaq.id — Website, Company Profile & Custom Systems