Selected Work

Projects & Case Studies.

GCP-native systems built for real operations — asset management, insurance intelligence, property data, and voice identity infrastructure.

IDist ®

Asset Management Infrastructure — Global Design & Procurement Platform

GCP Geospatial AI Asset Management Enterprise

4 Countries Unified Platform

30% Reduction Procurement Cycle

Enterprise Scale

View Case Study →

LH-PVI

Private Insurance — Mapping Unstructured Data to Geospatial Intelligence

Document AI Geospatial Insurance Data Pipeline

70% Faster Processing

Unstructured → Structured

Production ML

View Case Study →

NYCID

Private Property Dossier for NYC Residential

Real Estate AI Property Intelligence NYC GCP

8.1M Properties

Multi-Source Fusion

Verified Records

View Case Study →

Vera VB

Voice Biometric Identity Infrastructure

Voice AI Biometrics Privacy ML Pipeline

5 Layers Proof Stack

$0 Audio Stored on Platform

Private Beta

View Case Study →

Open Source

Technical Projects & OSS.

Publicly available implementations demonstrating production patterns for geospatial AI, RAG systems, and vendor intelligence.

Geospatial · AI

geoasset-verify-pipeline

github.com/danielflugger/geoasset-verify-pipeline

A GCP-native geospatial asset verification pipeline that ingests NYC property and infrastructure data, applies spatial reasoning and AI-assisted validation, and produces structured, auditable asset records. Built for real estate, insurance, and infrastructure operators who need verified, queryable spatial intelligence at scale.

What it does

Ingests raw municipal datasets and resolves spatial conflicts using PostGIS
Applies Gemini API for document extraction and field normalization
Outputs verified, structured asset records with full provenance chain
Deployed on Cloud Run with BigQuery as the queryable output layer

GCP PostGIS BigQuery Gemini API Cloud Run Python FastAPI

View on GitHub →

RAG · Document AI

rag-doc-processor

github.com/danielflugger/rag-doc-processor

A production-ready RAG (Retrieval-Augmented Generation) document processing system built on Vertex AI and GCP. Handles ingestion, chunking, embedding, retrieval, and generation for unstructured enterprise document collections — with observability, error handling, and output validation built in from day one, not bolted on after.

What it does

Multi-format document ingestion with intelligent chunking strategies
Vertex AI Embeddings + Vector Search for semantic retrieval
Pydantic output validation and structured response schemas
Full audit trail: query, retrieved context, and generated response

Vertex AI RAG Vector Search Pydantic Python FastAPI GCP

View on GitHub →

Governance · AI

vendor-mesh-intelligence

github.com/danielflugger/vendor-mesh-intelligence

An AI-assisted vendor dependency mapping and intelligence system. Analyzes organizational AI infrastructure to surface hidden dependencies, quantify vendor lock-in risk, and generate exit criteria for each critical dependency. Directly implements VERA's sovereignty patterns — the tool I built to solve my own problem and then open-sourced.

What it does

Ingests infrastructure configs and dependency graphs
Classifies vendors by criticality, replaceability, and data exposure
Generates structured exit criteria and migration cost estimates
Produces sovereignty risk reports in audit-ready format

AI Governance Vendor Analysis VERA Patterns Python TypeScript GCP

View on GitHub →

Projects & Case Studies.

Technical Projects & OSS.

More on GitHub.