
Meridian
Meridian is a personal executive intelligence platform that autonomously synthesizes your tasks, schedules, bills and documents into a unified dashboard & delivers clear, AI-driven daily morning briefings. It acts as your proactive digital assistant replacing chaotic app switching with a single pane of glass.
Taming the Chaos: How I Built an AI Powered Life Dashboard on AWS
Between managing work tasks, remembering to pay bills on time, keeping track of family events and trying to maintain healthy habits, life can quickly feel overwhelming. I found myself bouncing between five different apps just to figure out what my day looked like.
I realized I didn’t just need another to do list; I needed a personal executive assistant.
So for my latest hackathon project, I built Meridian, a centralized command center that unifies tasks, calendar schedules, financial obligations and health habits into a single pane of glass.
Here is a look at what I built and the AWS Serverless architecture that powers it.
What is Meridian?
Meridian is a web application that acts as your morning newspaper and daily planner combined. Instead of checking a calendar app, a finance app and a task manager you log into one beautiful interface that tells you exactly where you stand.
The best feature is the AI Morning Briefing. Every day, the application analyzes your pending tasks, upcoming bills, and calendar events, and generates a smart, natural language briefing. It tells you exactly what needs your attention right now, what you can defer, and congratulates you on the habits you're crushing.
The Architecture: Going Fully Serverless
From day one, I knew I wanted this project to be highly scalable, incredibly fast and cost effective. That meant going fully serverless on AWS.
I used the AWS Serverless Application Model (SAM) to define my infrastructure as code which allowed me to spin up and tear down my backend in seconds.
1. The Frontend: React & AWS Amplify
The user interface is a blazing fast React application built with Vite and Tailwind CSS. To get it into the hands of users, I deployed it using AWS Amplify Hosting. Amplify provided out of the box CI/CD, meaning every time I pushed code to my main branch on GitHub, it automatically built and deployed my frontend to a global CDN.
2. The Backend: API Gateway & AWS Lambda
All frontend requests are routed through Amazon API Gateway which triggers AWS Lambda functions running Node.js. By utilizing the modular AWS SDK for JavaScript (v3), I kept my Lambda deployment packages incredibly lightweight resulting in lightning fast cold starts.
3. The Database: DynamoDB Single Table Design
For storage, I chose Amazon DynamoDB. Instead of creating separate tables for Tasks, Bills, Events and Users, I implemented an advanced Single Table Design. By cleverly structuring my Partition Keys (PK) and Sort Keys (SK), I can fetch a user's entire dashboard state in a single sub millisecond database query.
4. Authentication: Amazon Cognito
Security is critical when dealing with personal life data. I integrated Amazon Cognito to handle all identity management. Using the Cognito Hosted UI, I wired up Google OAuth 2.0 allowing users to seamlessly sign in with their existing Google accounts and instantly receive secure JSON Web Tokens (JWTs) to access the API.
The Magic: Amazon Bedrock
The most exciting part of the project was building the AI Morning Briefing.
To do this, I utilized Amazon Bedrock. Every time a briefing is requested, a dedicated Lambda function gathers the user's uncompleted tasks and overdue bills from DynamoDB. It dynamically constructs a prompt and securely invokes the Anthropic Claude 3.5 Sonnet model via Bedrock.
The AI processes the context and streams back a personalized, highly relevant summary of the user's day. Because Bedrock is a fully managed service, I didn't have to worry about managing API keys or infrastructure for the LLM, it was entirely handled within my secure AWS environment via IAM roles.
Wrapping Up
Building Meridian was an incredible experience. It proved to me how powerful the AWS open source ecosystem is. Using tools like the AWS SAM CLI & the AWS SDK, I was able to rapidly build a highly complex enterprise grade architecture in a matter of days.
Not only did I solve a personal pain point by unifying my daily life tracking but I also got hands on experience with cutting edge generative AI through Amazon Bedrock.
I’m always looking to chat about serverless architecture, generative AI, or just geeking out over cool tech projects. If you found this build interesting, have feedback, or want to collaborate on something new, I’d love to connect!
You can reach out to me on LinkedIn (akshat-mohanty) or drop a comment below. Let’s build something awesome together!
Enjoyed reading this content? Let the author know!
Your likes, comments, shares, and saves help creators reach more builders.
Loading recommendations
Loading article