Astro Lab Meeting Maker — AI-Powered LinkedIn Outreach & Meeting Automation Platform
This project is a full-stack AI SaaS system that I designed and developed end-to-end, covering frontend, backend, system architecture, and AI-driven automation. The platform automates LinkedIn outreach workflows by managing lead ingestion, conversation handling, intent detection, and automated meeting booking through a single, scalable system.
From an engineering perspective, the focus of this project was building a reliable, production-ready automation platform capable of running thousands of parallel conversations, maintaining conversation state, and transitioning prospects through different stages of an outreach pipeline without manual intervention.
I architected the system to support multi-agent execution, real-time message processing, pipeline state tracking, analytics, and external service integrations while ensuring scalability, fault tolerance, and clean separation between business logic, AI decision layers, and communication workflows.
- Designing the overall system architecture for an AI-driven outreach SaaS
- Building the frontend dashboard for managing agents, leads, pipelines, and analytics
- Developing the backend services and APIs to handle outreach workflows, lead states, and conversations
- Implementing AI agent orchestration logic for message generation, reply handling, and intent-based decision making
- Creating a conversation state engine to track lead progress from first contact to meeting booked
- Building multi-tenant support for multiple accounts and agents
- Implementing analytics, logging, and monitoring for lead journeys and system performance
- Optimizing for scalability, concurrency, and reliability in a production environment
Category
Full Stack Development
Roles
Full Stack Development - Next.js and Supabase
Location
Melbourne, Australia
Published
September 25, 2024
Check out Some Screenshorts
(8/30) Screen Shorts are displayed, for complete demo checkout my demo on linkedin
Challenges Faced & Final Results
One of the main challenges was designing a backend system capable of handling long-running, stateful outreach workflows, including asynchronous replies, follow-ups, intent changes, and meeting bookings, while keeping lead states consistent and reliable. Additional challenges included integrating multiple third-party services with proper error handling, rate-limit protection, and retries, as well as separating AI decision logic from core business workflows to keep the system maintainable and debuggable.
The final result is a production-ready, scalable full-stack SaaS platform that reliably manages large volumes of concurrent conversations, tracks each lead’s journey accurately, and automates complex outreach workflows end-to-end. This project demonstrates my ability to architect, build, and ship robust systems that operate under real-world constraints and scale beyond simple CRUD applications.
