Back to Projects

Project

Investify

Market Intelligence & ML Prediction Platform

A full-stack investing platform with a Go API, FastAPI ML service, Next.js dashboard, technical feature pipelines, and daily stock prediction workflows.

GoPostgreSQLFastAPIPythonscikit-learnNext.jsReactDocker

Overview

Investify is a full-stack market intelligence platform built around ticker discovery, historical price ingestion, technical feature generation, and ML-backed predictions. It separates the production API, model service, and web dashboard so market data workflows can run independently from the user-facing experience.

Investify Architecture

Investify connects a Next.js dashboard to a Go API, Postgres data layer, and FastAPI ML service for daily prediction and portfolio intelligence workflows.

web

Next.js Dashboard

Ticker browsing, holdings management, charts, confidence gauges, and admin controls.

api

Go API

Chi router with auth, holdings, ticker, price, feature, and admin endpoints.

market

Market Data Providers

Twelve Data integration with a synthetic development provider fallback.

features

Feature Pipeline

Backfills historical prices and computes SMA, EMA, RSI, MACD, momentum, and volatility features.

ml

FastAPI ML Service

Prediction and training routes powered by pandas, scikit-learn, and stored model artifacts.

db

PostgreSQL

Stores users, tickers, holdings, prices, features, credentials, and pipeline jobs.

Service / Data Flow

Next.js DashboardGo API
Dashboard requests
Go APIMarket Data Providers
Price ingestion
Market Data ProvidersFeature Pipeline
Historical prices
Feature PipelinePostgreSQL
Technical indicators
Go APIFastAPI ML Service
Prediction / train calls
FastAPI ML ServicePostgreSQL
Training data and predictions
Go APIPostgreSQL
Auth, holdings, tickers

What I built

  • Go backend with JWT auth, role-aware admin routes, holdings CRUD, ticker management, market data ingestion, and encrypted provider credentials
  • FastAPI ML service with model training jobs, prediction endpoints, rule-based fallback logic, and stored scikit-learn artifacts
  • Technical analysis pipeline for SMA, EMA, RSI, MACD, momentum, volatility, historical prices, and feature snapshots
  • Next.js dashboard for active tickers, ticker detail pages, holdings, training controls, price charts, confidence gauges, and feature visibility
  • Postgres-backed migrations covering users, API credentials, ticker metadata, constraints, roles, and pipeline jobs
  • Dockerized service boundaries for the API, ML service, and web app