PROJECTS

Enterprise scope.
Public evidence.

Selected work across enterprise data engineering and public data products. Enterprise descriptions are sanitized; public projects link to inspectable code.

ENTERPRISE WORK

Professional experience

Enterprise lakehouse architecture

Taking a Databricks platform from proof of concept to production, with integration patterns, modeling standards, and governance used across engineering and analytics teams.

Technical architecture and platform engineering

DatabricksSparkDelta LakeUnity CatalogAWS

ENTERPRISE WORK

Professional experience

Applied AI over enterprise data

LLM and RAG architecture for document processing, structured extraction, metadata enrichment, vector search, and enterprise search over governed data.

LLM/RAG workflow architecture

LLMsRAGStructured extractionVector search

ENTERPRISE WORK

Ongoing initiative

Enterprise CRM migration

Data engineering technical ownership across migration, integration architecture, reconciliation, reporting, downstream modeling, and governance.

Technical owner, data engineering

PUBLIC PROJECT

Public implementation

DataNepal

A public data product with canonical geography, required provenance, and a clear boundary between ingestion, transformation, and publication.

Public engineering project

PythondltDuckDBdbtParquetCloudflare

EARLIER PUBLIC WORK

Archived

Voter aggregate insights

An earlier application using DuckDB to prepare aggregates, ClickHouse for analytical queries, and FastAPI to expose province-to-ward statistics.

DuckDBClickHouseFastAPIDocker