For most of my career, I worked on large-scale commerce search systems powered by Oracle Endeca.
Search wasn’t just a feature. It impacted revenue, conversions, and user experience at scale.
Recently, I started building hands-on projects with Elasticsearch.
One thing became clear very quickly:
The tools have changed. The fundamentals haven’t.
What Years in Search Really Teach You
Working on enterprise search platforms forces you to think about:
Index design for large catalogs
Relevance tuning and ranking strategy
Faceted navigation
Query performance under load
Data modeling for search behavior
Those principles are not product-specific.
They apply just as much to Elasticsearch.
What I Built
To validate the transition, I implemented a small commerce-style Elasticsearch project that included:
Explicit index mappings (separating text and keyword fields)
Weighted relevance boosting (title^3 vs description)
Bulk indexing to simulate production ingestion
Aggregations for faceted navigation
Instead of focusing on syntax, I focused on architecture:
How should the index be structured?
How should ranking reflect business logic?
How would this behave at scale?
That thinking felt very familiar.
What I’m Working on Now
I’m expanding the project to simulate more production concerns:
Larger datasets
Pagination strategy
Index optimization
Performance behavior
The goal isn’t just learning a new tool.
It’s applying deep search engineering principles to modern platforms like Elasticsearch and OpenSearch.
Technology evolves.
But strong system design thinking compounds.
If you're modernizing search, migrating from legacy platforms, or tuning relevance in high-volume systems, I’d be happy to connect and exchange ideas.
You can explore the project code on GitHubhere .