Home

Portfolio

AI Home Management Platform

Home

Portfolio

AI Home Management Platform

Home

Portfolio

AI Home Management Platform

AI Property Management Platform

Industry: Real Estate / PropTech / AI


Services: Solution & Architecture Audit, Full Platform Rebuilding, Mobile App Development, UI/UX Audit, UI/UX Design


Tech Stack: .Typescript, Node.js, Postgres, AWS Serverless, Amazon Bedrock, Textract, OpenAI, LangChain, Swift


Team: 5 specialists

About the client

About the client

Cassi began as a vision to simplify the everyday chaos of homeownership — turning documents, insurance, maintenance, and service coordination into something effortless. Built around a chat-first experience, the platform uses LLMs, a household knowledge graph, and automated workflows to anticipate needs before they become tasks.

Cassi began as a vision to simplify the everyday chaos of homeownership — turning documents, insurance, maintenance, and service coordination into something effortless. Built around a chat-first experience, the platform uses LLMs, a household knowledge graph, and automated workflows to anticipate needs before they become tasks.

Cassi began as a vision to simplify the everyday chaos of homeownership — turning documents, insurance, maintenance, and service coordination into something effortless. Built around a chat-first experience, the platform uses LLMs, a household knowledge graph, and automated workflows to anticipate needs before they become tasks.

When the founders came to ExergyIQ, the initial version built by a in-house team wasn’t matching that vision. After a full product and architecture audit, we rebuilt the solution from the ground up, creating two fully redesigned mobile apps for homeowners and service providers — this time aligned with Cassi’s mission to become the intelligent home assistant every household can rely on.

When the founders came to ExergyIQ, the initial version built by a in-house team wasn’t matching that vision. After a full product and architecture audit, we rebuilt the solution from the ground up, creating two fully redesigned mobile apps for homeowners and service providers — this time aligned with Cassi’s mission to become the intelligent home assistant every household can rely on.

When the founders came to ExergyIQ, the initial version built by a in-house team wasn’t matching that vision. After a full product and architecture audit, we rebuilt the solution from the ground up, creating two fully redesigned mobile apps for homeowners and service providers— this time aligned with Cassi’s mission to become the intelligent home assistant every household can rely on.

results

File ingestion success rate increased from 65% to 98%

File ingestion success rate increased from 65% to 98%

Scan

Upload

Email

Auto-classification accuracy improved by 30%

Auto-classification accuracy improved by 30%

Reduced operational complexity for the client team

Reduced operational complexity for the client team

Faster onboarding for homeowners and service providers

Faster onboarding for homeowners and service providers

Time-to-process each uploaded document reduced by 55%

Time-to-process each uploaded document reduced by 55%

Saving in memory...

GOALS

Replace the low-quality prototype built by an in-house team

Replace the low-quality prototype built by an in-house team

Replace the low-quality prototype built by an in-house team

Establish a scalable, enterprise-ready architecture

Establish a scalable, enterprise-ready architecture

Establish a scalable, enterprise-ready architecture

Rebuild two new mobile applications from scratch using modern tech stack

Rebuild two new mobile applications from scratch using modern tech stack

Rebuild two new mobile applications from scratch using modern tech stack

Enable automated document ingestion (warranties, insurance, receipts, manuals)

Enable automated document ingestion (warranties, insurance, receipts, manuals)

Enable automated document ingestion (warranties, insurance, receipts, manuals)

Architecture review and consulting on a future AI agent implementation

Architecture review and consulting on a future AI agent implementation

Architecture review and consulting on a future AI agent implementation

Ensure strong security foundations given sensitive user data

Ensure strong security foundations given sensitive user data

Ensure strong security foundations given sensitive user data

Protection

Maintenance

Coordination

Financials

Organization

Key Challenges

The platform suffered from multiple issues affecting both business and engineering:

The platform suffered from multiple issues affecting both business and engineering:

Poor code quality and unstable architecture inherited from the in-house team

Poor code quality and unstable architecture inherited from the in-house team

Unscalable backend with no clear boundaries or workflows

Unscalable backend with no clear boundaries or workflows

Inconsistencies between product vision and existing implementation

Lack of engineering standards and code review discipline

No proper orchestration for document ingestion or AI use cases

No proper orchestration for document ingestion or AI use cases

High risk related to handling sensitive insurance and home documents

Inefficient manual processes (especially claim matching)

Legacy ingestion flow not designed for scale, making it difficult to support additional AI features

Legacy ingestion flow not designed for scale, making it difficult to support additional AI features

Lack of UX alignment across both apps

Inefficient manual processes (especially claim matching)

Missing documentation and unclear product flows

Missing documentation and unclear product flows

OUR APPROACH

01

Discovery & Architecture Audit (1.5 months)

• Conducted a deep analysis of the previous solution

• Identified architectural flaws, security risks, and scalability issues

• Defined a new system architecture based on AWS Serverless

• Designed a clean separation for two applications: Homeowner & Service Provider

• Mapped all AI workflows, knowledge graph logic, and document ingestion pipelines

02

Rebuilding the Platform from Scratch

• Re-architected backend to support modular growth

• Implemented secure document ingestion with AWS Managed services Textract

• Implemented event-driven orchestration for automated tasks

• Set foundations for real-time insights and recommendations

03

AI Agent & Automation Layer

• Designed conversational UX with a chat-first interaction model

• Created reusable components to streamline future AI feature development

• Set the Architecture for a future AI agent implementation

• Established RAG (Retrieval-Augmented Generation) pipelines using household data

04

Mobile Applications Development

Homeowner App (iOS)

Refactored 10+ backend repositories

Refactored 10+ backend repositories

• AI-powered home knowledge base
• Document scanning and automatic classification
• Personalized insights and alerts
• Task and maintenance management

Service Provider App

• Job intake and scheduling
• Communication with homeowners
• Task tracking
• Document attachments and updates

05

Quality Assurance & Delivery Management

• Introduced structured QA processes

• Continuous testing for AI outputs and document processing

• Weekly product alignment with the founders

• Release planning optimized for investor milestones

What we achieved

A rebuilt, scalable platform

A rebuilt, scalable platform

A rebuilt, scalable platform

AI-powered home knowledge base

AI-powered home knowledge base

AI-powered home knowledge base

End-to-end automation

End-to-end automation

End-to-end automation

High-quality mobile applications

High-quality mobile applications

High-quality mobile applications

Long-term scalability unlocked

Long-term scalability unlocked

Long-term scalability unlocked

Streamlined product operations

Streamlined product operations

Streamlined product operations

Team

Solution Architect & Backend Engineer

Solution Architect &

Backend Engineer

iOS Developer

UI/UX Designer

Delivery Manager

QA Manual Engineer

Contact

Got a similar challenge with your platform?

Got a similar challenge with your platform?

Got a similar challenge with your platform?

Let’s discuss architecture, modernization options, automation, or AI integration.