All work
in progress2026 · Solo full-stack + AI engineer
Lagos Property Intelligence
An AI-assisted investment platform for reading a property market, not just browsing listings.
01 / The problem
Property investment decisions in fast-moving markets like Lagos are usually made on gut feeling and word of mouth, because the data that would support a real valuation is scattered, informal, or not digitized at all.
02 / The architecture
Structured a pipeline that ingests listing and transaction data, normalizes it into a consistent valuation schema, and layers an AI analysis stage on top that explains — not just predicts — why a property is priced where it is relative to its neighborhood trend.
03 / The challenges
Real estate data in this market is inconsistent by nature — addresses, currency formats, and listing quality vary wildly. Most of the actual engineering effort went into a resilient normalization layer, not the AI itself.
04 / The results
A working prototype that turns fragmented listing data into a valuation view a buyer can actually reason about, with trend context instead of a bare price tag.
05 / The lessons
The hard part of applied AI is almost never the model call — it's building data pipelines trustworthy enough that the model's output means something.
06 / Metrics
1,000+
Listings normalized
3
Data sources unified
07 / Tech stack
- Next.js
- TypeScript
- Python
- OpenAI
- PostgreSQL
- Azure
Contact
Let's Build Something Worth Remembering.
Open to internships, founding-engineer roles, and focused contract work in cloud, data, and AI products.