🚀 Inspiration
In today’s world, people use multiple platforms like Google Calendar, Gmail, GitHub, and Slack daily. Managing all these tools separately can be time-consuming and inefficient.
We were inspired to build a single AI assistant that can understand natural language and automate tasks across multiple services in one place.
💡 What it does
CipherMate AI is an intelligent assistant that allows users to perform real-world actions using simple commands like:
- “Schedule a meeting tomorrow at 3 PM”
- “Send an email to my team”
- “Create a GitHub issue”
- “Send a Slack message”
The AI understands the request, routes it to the correct service, and executes the task seamlessly.
⚙️ How we built it
We designed CipherMate using a modular AI-driven architecture:
- Frontend: Next.js + Tailwind CSS
- Backend: FastAPI (Python)
- AI Layer: OpenAI Agents SDK (with Gemini API)
- Authentication: Auth0 (OAuth-based secure access)
- Integrations:
- Google Calendar API
- Gmail API
- GitHub API
- Slack API
- Google Calendar API
We implemented a triage agent system that routes user requests to specialized agents like calendar, email, GitHub, and Slack.
🧩 Challenges we ran into
- Handling OAuth permissions across multiple services
- Managing API integrations and tokens securely
- Reducing AI response delays
- Designing a scalable and extensible architecture
🏆 Accomplishments that we're proud of
- Built a fully working multi-service AI assistant
- Integrated 4 major platforms in one system
- Created a clean and extensible architecture
- Enabled real-time task execution using natural language
📚 What we learned
- How to design AI-driven systems
- Working with OAuth and secure authentication (Auth0)
- Integrating multiple third-party APIs
- Building scalable backend architecture
🔮 What's next for CipherMate AI
- Add more integrations (Notion, Trello, WhatsApp)
- Improve AI accuracy and speed
- Add voice command support
- Build mobile app version
Log in or sign up for Devpost to join the conversation.