Real Estate CRM Platform
This Ospina case study documents how Carlos Rico-Ospina approached a specific risk, infrastructure, revenue, or research problem and what was built to address it.
Built an intelligent parsing system that turns messy broker blasts into structured listings and auto-matches them to client criteria.
The Problem
Broker blasts arrive in inconsistent formats (Excel → PDF), making them painful to operationalize. The bottleneck isn't the CRM UI—it's ingestion and normalization. If you can reliably parse the blast, everything downstream becomes automatable.
The Insight
Each property record sits between two address anchors. By detecting addresses and inferring boundaries, you can construct structured property objects from chaotic input. Combine deterministic parsing with AI enrichment where it helps, rules where correctness matters.
What I Built
- Built parsing system accepting multiple spreadsheet formats and messy exports
- Implemented address-boundary inference to detect property record boundaries
- Constructed structured property objects with AI-assisted enrichment
- Created auto-matching engine to notify clients when listings match their criteria
Outcomes
- Converted recurring manual task into scalable ingest → normalize → enrich → match → notify workflow
- Handles chaotic upstream data through pragmatic 'real-world parsing' techniques
- Demonstrates safe combination of deterministic heuristics with AI enrichment
Have a Similar Challenge?
Let's discuss how I can help you achieve similar results.
Start a Conversation