AI-Powered Housing Valuation Web App
Budget: $10 – $30 CAD
THIS IS FOR CONTINUOUS WORK as this project is going to develop into a business and the business is already registered. I am looking for SALARIED WORK for a long term contract.
I am a software engineer myself.
I’m building an interactive web application that can estimate housing prices on the fly by talking to users through a large-language-model–driven interface. The core of the project is an AI module that combines an LLM with Retrieval-Augmented Generation (RAG) so the system can reason over multiple data sources and justify every figure it returns.
The model must pull insights from:
• Historical sales data
• Current market listings
• Key economic indicators
• Real-time information the user types in
• PDFs of current listings fed into the RAG pipeline
Workflow I have in mind
1. A clean front-end where users upload a PDF or enter an address, tweak assumptions, and chat with the AI for clarifications.
2. A back-end that calls the LLM, performs retrieval over the structured databases and the uploaded documents, fuses the results, then serves a price estimate plus an explanation trail.
3. An admin view that lets me update data sources and monitor API usage.
Deliverables
• End-to-end web app (front-end + back-end) deployed to a cloud instance of my choice
• RAG pipeline wired to the LLM and the listed datasets/PDFs
• Unit tests or notebooks proving the estimation logic works on sample properties
• README with setup, environment variables, and fine-tuning instructions
If you already have experience integrating OpenAI, LangChain, Pinecone, Vector DBs, or similar tooling, that’s perfect—but I’m open to whichever stack you feel is most reliable. Let me know how you would architect the data layer and the chat interface and how long you’ll need to ship a demo.
I am a software engineer myself.
I’m building an interactive web application that can estimate housing prices on the fly by talking to users through a large-language-model–driven interface. The core of the project is an AI module that combines an LLM with Retrieval-Augmented Generation (RAG) so the system can reason over multiple data sources and justify every figure it returns.
The model must pull insights from:
• Historical sales data
• Current market listings
• Key economic indicators
• Real-time information the user types in
• PDFs of current listings fed into the RAG pipeline
Workflow I have in mind
1. A clean front-end where users upload a PDF or enter an address, tweak assumptions, and chat with the AI for clarifications.
2. A back-end that calls the LLM, performs retrieval over the structured databases and the uploaded documents, fuses the results, then serves a price estimate plus an explanation trail.
3. An admin view that lets me update data sources and monitor API usage.
Deliverables
• End-to-end web app (front-end + back-end) deployed to a cloud instance of my choice
• RAG pipeline wired to the LLM and the listed datasets/PDFs
• Unit tests or notebooks proving the estimation logic works on sample properties
• README with setup, environment variables, and fine-tuning instructions
If you already have experience integrating OpenAI, LangChain, Pinecone, Vector DBs, or similar tooling, that’s perfect—but I’m open to whichever stack you feel is most reliable. Let me know how you would architect the data layer and the chat interface and how long you’ll need to ship a demo.