Tommy
van der Burg
Building systems that create real business value.
From idea to production.
Tommy
van der Burg
Building systems that create real business value.
From idea to production.
Career
The journey so far
From infrastructure to data to AI — each step sharpened a different layer of the stack.
AI Engineer
CurrentData Engineer
DevOps Engineer
Business IT & Management
EducationWork
Featured projects
Production systems, not proofs of concept.
Reusable AI Component SDKs
Stop reinventing the wheel. Build once, used by all.
Problem
Every AI team at NN was building its own integrations from scratch — each with their own way of calling an LLM, handling speech, or connecting to internal systems. No consistency, no governance, duplicated effort across the board.
Solution
Built a set of generic SDKs that any team at NN can drop into their solution: a standardised LLM caller, an STT module, a PII masking interface, and NN-specific connectors for Gaia (the internal RAG solution) and SharePoint.
Automated Call Logging
10,000 calls processed daily, zero manual effort
Problem
Call agents manually logged a summary and classified the reason for every call — a slow, inconsistent process that didn't scale across thousands of interactions a day.
Solution
End-to-end AI pipeline using Azure OpenAI for summarisation and call reason classification. Outputs pushed directly into Salesforce and other CRM systems used across NN. Full CI/CD deployed system with Terraform infrastructure, logging, security, and monitoring in place.
CCDAP — Customer Data Platform
NN's central data lakehouse for customer data
Problem
The legacy data warehouse couldn't keep pace with the organisation. Business units ran their own pipelines with no shared governance, making reliable data a constant negotiation.
Solution
Co-built CCDAP — a medallion architecture lakehouse (Bronze → Silver → Gold) on Databricks, holding all NN customer data and serving multiple business units. ETL flows in Azure Data Factory; ingestion and materialisation in Databricks notebooks. Delivered governance modules for data quality, lineage, and access control.
Automated Call Logging
10,000 calls processed daily, zero manual effort
Problem
Call agents manually logged a summary and classified the reason for every call — a slow, inconsistent process that didn't scale across thousands of interactions a day.
Solution
End-to-end AI pipeline using Azure OpenAI for summarisation and call reason classification. Outputs pushed directly into Salesforce and other CRM systems used across NN. Full CI/CD deployed system with Terraform infrastructure, logging, security, and monitoring in place.
Stack
Skills & tools
Tools chosen for production reliability, not resume padding.
AI & ML
Cloud & Infra
Data
Engineering
Business & Strategy
Human
Outside of working hours
A well-balanced lifestyle enhances creativity and keeps the thinking sharp.