FutureSelf AI
Inspiration
Every day, people make decisions they later regret.
We buy products we never use, send emails we wish we could take back, accept opportunities without enough research, and make important choices while influenced by emotions or incomplete information. Most digital tools help people act faster, but very few help people make better decisions.
The idea for FutureSelf AI came from a simple question:
What if you could ask your future self for advice before making an important decision?
Instead of helping users do more, we wanted to help them regret less. Our goal was to build an AI-powered decision intelligence system that acts as a layer between human impulses and online actions.
What It Does
FutureSelf AI is a Chrome extension that analyzes important online decisions before they happen.
When a user is about to:
- Purchase a product
- Send an email
- Apply for a job
- Subscribe to a service
- Make a financial decision
the extension intercepts the action and generates a Future Impact Report.
The system evaluates the decision using multiple specialized AI agents and provides insights such as:
- Financial impact
- Opportunity cost
- Emotional risk
- Long-term consequences
- Personalized regret probability
Instead of simply answering questions, FutureSelf AI simulates possible future outcomes and helps users make more informed choices.
How We Built It
The project combines browser automation, AI agents, memory systems, and modern web technologies.
Frontend
- React
- TypeScript
- Tailwind CSS
- Framer Motion
Chrome Extension
- Manifest V3
- Content Scripts
- Background Service Worker
- Chrome Storage API
Backend
- FastAPI / Next.js API
- Multi-Agent Orchestration Layer
- OpenAI API
Data Layer
- PostgreSQL
- ChromaDB Vector Database
AI Agent System
FutureSelf AI uses multiple agents working together:
Financial Agent
Analyzes affordability and financial impact.
Opportunity Cost Agent
Calculates alternative uses of money and future value.
Risk Agent
Evaluates uncertainty and downside scenarios.
Emotion Agent
Detects emotionally charged language in emails and messages.
Memory Agent
Learns from previous decisions and generates personalized recommendations.
The outputs of all agents are aggregated into a final decision score and recommendation.
Challenges We Faced
One of the biggest challenges was designing a system that could analyze completely different types of decisions while maintaining a consistent user experience.
A purchase decision requires financial reasoning, while an email requires emotional analysis, and a job application requires career evaluation. Building a flexible multi-agent architecture capable of handling all these scenarios was a significant design challenge.
Another challenge was balancing helpful guidance with user autonomy. We wanted FutureSelf AI to provide valuable insights without becoming intrusive or preventing users from making their own choices.
Creating a memory layer was also challenging because recommendations become significantly more useful when they are personalized. We needed a way to store historical decisions and retrieve relevant context efficiently for future analyses.
What We Learned
Through this project we learned:
- Multi-agent AI system design
- Browser extension architecture
- Decision intelligence workflows
- Vector databases and memory systems
- Context-aware AI recommendations
- Real-time user interaction design
- AI orchestration and agent communication
Most importantly, we learned that AI can be used for more than productivity. It can help people make better decisions by providing perspective before actions are taken.
Future Vision
We believe FutureSelf AI can evolve into a personal decision operating system.
Future versions could include:
- Investment decision analysis
- Relationship and communication guidance
- Life planning simulations
- Financial forecasting
- Personalized decision coaching
- Future-self scenario modeling
Our long-term vision is simple:
Help people make decisions today that their future selves will thank them for.
Impact
FutureSelf AI shifts AI from being a tool that accelerates actions to a system that improves judgment.
Instead of helping users move faster, it helps them move smarter.
Because sometimes the best decision is not the one that is easiest to make—it is the one you will be happy you made years later.
Built With
- background-service-worker
- chromadb
- chrome-extension-api
- chrome-storage-api
- clerk
- content-scripts
- decision-intelligence-system
- docker
- fastapi
- framer-motion
- git
- github
- javascript
- jwt-authentication
- langgraph
- manifest-v3
- multi-agent-ai
- next.js
- openai-api
- postgresql
- python
- railway
- react
- rest-api
- retrieval-augmented-generation
- tailwind-css
- typescript
- vector-embeddings
- vercel
Log in or sign up for Devpost to join the conversation.