Mirai Analytics is an AI-powered financial assistant and analytics platform designed for private. It combines real-time market data, news analysis, portfolio insights, and conversational AI to help users make smarter investment decisions.
-
AI Investment Voice Assistant
Chat with an AI that understands market trends, portfolio impacts, and can answer investment questions in natural language. -
Portfolio Analytics
Visualize your holdings, sector allocations, and receive actionable rebalancing suggestions. Get friendly, practical recommendations based on your risk profile. -
Market Overview
See real-time market data, trends, and sector performance in a clean dashboard. -
News & Insights
Aggregates and analyzes financial news from top sources (Reuters, Bloomberg, FT, Yahoo Finance, CNBC).
Highlights important articles, scores impact, and provides concise summaries and actionable insights. -
Onboarding Flow
Personalized onboarding to capture user risk tolerance, sector interests, and notification preferences. -
AI-Generated Tips & Insights
Get premium AI insights and tips tailored to your portfolio and interests. -
Realtime & Scheduled Data Ingestion
Automated backend pipelines scrape, normalize, and analyze news and market data.
Frontend:
- React Native (Expo)
- TypeScript
- NativeWind (Tailwind CSS for React Native)
- Lucide Icons
- React Navigation
- TanStack React Query
Backend:
- Python 3.12
- Quart (async Flask-like web framework)
- SQLAlchemy (async ORM)
- PostgreSQL
- LangChain & LangGraph (AI workflow orchestration)
- Anthropic API (LLMs for extraction, scoring, and writing)
- OpenAI API
- Twilio (voice assistant integration)
- ElevenLabs (voice synthesis)
- Async web scraping (Crawl4AI, Playwright)
Other:
- Pydantic (schemas & validation)
- Docker/Conda (environment management)
- Task scheduling (quart-tasks)
- Safe area/context for mobile UI
-
Frontend:
Mobile-first, modular React Native app with screens for Home, News, Portfolio, and Assistant.
Uses hooks and context for state management and onboarding persistence. -
Backend:
Async API endpoints for news, insights, portfolio analysis, and recommendations.
Modular pipelines for scraping, normalization, fact extraction, impact scoring, and verification.
LLM-powered summarization and conversational responses. -
Data Flow:
- Scrape news → Normalize → Extract facts → Score impact → Write analyst brief → Verify output
- Portfolio data → Analyze allocation, risk, and trends → Generate recommendations