LombardGPT

LombardGPT project image.

Summary

AI-powered market network for commercial real estate companies. The platform uses predictive AI trained on CRE company data to help verified buyers and sellers match against specific properties.

My Role

Lead Software Engineer responsible for full-stack architecture, product delivery, integrations, deployment, and core platform features.

Key Outcomes

  • Led development of a PropTech SaaS platform used by 20K+ active users.
  • Improved engagement by 70% through UI performance optimization and feature rollout.
  • Shipped real-time chat, AI content streaming, subscriptions, organization invites, and file/calendar integrations.
  • Architected AWS deployment across Amplify, EC2, Lambda, and DynamoDB.

Stack

Next.js, React, Tailwind CSS, Headless UI, AWS, PostgreSQL, DynamoDB, Prisma, Soketi, Auth0, Stripe, next-i18next.


Delivery Details

LombardGPT is an AI-powered Market Network for Commercial Real Estate companies that uses predictive Artificial Intelligence algorithm trained on CRE companies' data points and helps verified buyers and sellers to match against a specific property.

Website


Role

Lead Software Engineer.

Technologies Used:

My Contributions:

Full-Stack Web Application Development:

  • Developed the entire web application from scratch using Next.js 13.

Database Schema Design:

  • Designed a robust and scalable database schema in PostgreSQL using Prisma ORM to manage complex relationships and ensure data integrity.

Authentication Integration:

  • Integrated Auth0 for secure user authentication, enabling users to log in using social accounts and standard email/password credentials.

Calendar and Drive Integration:

  • Integrated Google and Outlook calendars using their respective APIs to synchronize user events within the app.
  • Enabled Google Drive and OneDrive integration for seamless file selection and usage within the CRM application.

AI Content Streaming:

  • Integrated live text streaming for AI-generated content such as summaries and emails, fetching data from Django API.

Real-Time Chat Feature:

  • Developed a comprehensive chat feature allowing buyers and sellers to engage in live conversations, share files, and create group chats.
  • Hosted a Soketi socket server on an EC2 instance to facilitate real-time communication.

Queue Management for Delayed Tasks:

  • Implemented a custom queue using AWS SQS and AWS Lambda to handle time-consuming tasks, improving overall response times.

Payment Integration:

  • Integrated Stripe for subscription management, enabling users to purchase plans and access features based on their subscription level.

Member Invite Management:

  • Developed features for admin users to invite members to their organization and use organization's subscription collectively.

Customizable Themes:

  • Implemented a theme customization feature using CSS variables, allowing users to change the app’s appearance based on their preferences.

Design and Dynamic Components:

  • Utilized Tailwind CSS and Headless UI library for creating responsive and dynamic components, ensuring a seamless user experience.

Internationalization:

  • Integrated internationalization using the next-i18next package to support multiple languages, making the app accessible to a global audience.