Jonathan Hazeley
Principal Data & AI Architect
Platforms fail on trust, not on code. I architect Databricks lakehouse and AI systems, and prove them before a business commits.
Charlotte, North Carolina
- rows migrated and proven on one pilot
- 678M
- automated checks per validation run
- 114
- reusable Databricks skills in the accelerator
- 118
- Microsoft and Databricks ECIF proposal I wrote the case for
- ~$1.5MM
rows migrated and proven on one pilot
automated checks per validation run
reusable Databricks skills in the accelerator
Microsoft and Databricks ECIF proposal I wrote the case for
The evidence
Selected work
A few problems worth explaining in full.
Proving a Migration, Not Just Doing One
Ten source tables moving platforms. No way to prove any of them arrived intact.
- rows migrated on the pilot
- 678M
- structural match on every validation run
- 100%
- automated checks per run
- 114
- average difference against source (2% tolerance)
- 0.01%
rows migrated on the pilot
structural match on every validation run
automated checks per run
average difference against source (2% tolerance)
- Databricks
- Lakeflow Declarative Pipelines
- Lakehouse Federation
- Snowflake
- Delta Lake (MERGE, Liquid Clustering)
- Medallion architecture
The Databricks Catalog Other Engineers Build Inside
Every engagement re-derived the same decisions. The knowledge lived in whoever was staffed.
- reusable skills, client-generic by construction
- 118
- categories, from platform core to AI and agents
- 13
reusable skills, client-generic by construction
categories, from platform core to AI and agents
- Databricks
- Python
- FastMCP
- Model Context Protocol
Architecture Enforced by Gates, Not Convention
Architecture documents describe intent. Codebases drift from it within months.
- packages split on a network-and-secrets axis
- 7
- AST gates failing CI on a boundary violation
- 4
packages split on a network-and-secrets axis
AST gates failing CI on a boundary violation
- Python
- PostgreSQL
- Databricks
- AST analysis
Statewide Transportation Data, Made Trustworthy
A feed lapsed. The dashboard still rendered, with a gap nobody could see.
- improvement in statewide reporting availability
- 30%
- faster onboarding for a new data source
- 60%
improvement in statewide reporting availability
faster onboarding for a new data source
- Databricks
- Delta Live Tables
- PySpark
- Azure OpenAI
The Metrics Behind a 50% EBITA Increase
Everyone had the data. Nobody had a reason to believe it.
- EBITA increase (~$50M) the pipelines contributed to
- 50%
- Databricks pipelines integrating IoT and SaaS sources
- 15+
EBITA increase (~$50M) the pipelines contributed to
Databricks pipelines integrating IoT and SaaS sources
- Databricks
- SQL
- Python
- Power BI
The method
How I run an engagement
What happens on all of them, not what happened on one. Every phase names the artifact that leaves it.
- Assess
Name the problem beneath the stated problem
The problem a team reports is rarely the one worth solving. Before any architecture, I establish what breaks if the numbers are wrong, and who has to defend them. That decides what the work has to prove, not only what it has to build.
Ships
The baseline, and the definition of done
- Design
Build the proof with the system
Validation is designed alongside the thing it validates, not bolted on after it. The checks are config-driven, so a new table is one line of config and a new skill is a graded eval, never new rule code. The standard does not decay as the surface grows.
Ships
Target architecture and the validation contract
- Build
Ship a domain at a time
Delivery runs domain by domain. On a migration that means a parallel run against the source, with every table proven equivalent before anything is retired. A framework that only ever confirms what you hoped is not a validation framework, so the comparison has to be able to fail.
Ships
Production systems, and a passing check on each
- Hand over
Make the standard outlive me
A gate nobody runs is indistinguishable from no gate at all. The checks run in CI, not on request. The definitions live in a glossary the team owns, not in one engineer’s head.
Ships
Runbook, glossary, and the checks in CI
Where I am
Currently
Lead Data Engineer Consultant at Lovelytics. Databricks’ four-year Partner of the Year, backed by Databricks Ventures. I am the client-facing technical lead on two to three concurrent lakehouse builds and migrations, and I set the Databricks delivery standard the engagement teams work inside.
Full experienceVerified
Credentials
Certified on the platforms I build on. Every one links to its issuer.
- Databricks Certified Data Engineer Associate
- AWS Certified Cloud Practitioner
Plus 13 training and course credentials across Data Engineering, AI, Cloud & Delivery.
The toolkit
What I work with
- Languages & processing
- SQL
- Python
- PySpark
- Apache Spark (Structured & Streaming)
- Working knowledge:
- Pandas
- NumPy
- Databricks & lakehouse architecture
- Databricks
- Unity Catalog governance
- Lakeflow Declarative Pipelines
- Delta Live Tables
- Delta Lake
- Databricks Asset Bundles
- Auto Loader
- Medallion architecture (bronze/silver/gold)
- Data Vault 2.0
- Data Mesh
- Dimensional modeling (SCD Type 1/2)
- Migration architecture
- Storage & query engines
- Snowflake
- BigQuery
- Postgres
- Working knowledge:
- Amazon Athena
- Presto
- Trino
- Cloud & delivery
- Azure (ADLS, Fabric)
- AWS
- GCP
- ETL / ELT
- CI/CD for data
- Environment promotion
- Azure DevOps
- GitHub Actions
- Git
- dbt
- Fivetran
- Astronomer
- Working knowledge:
- Apache Airflow
- AI & agentic systems
- Azure OpenAI
- Model Context Protocol
- Claude Agent SDK
- Prompt design & engineering
- Agent evaluation
- Analytics & BI
- Power BI
- Power BI semantic modeling (DAX)
- Epic Clarity / Caboodle
- Heap.io
- Celonis
- Working knowledge:
- Tableau
- Redash
- Practice & leadership
- Databricks migrations
- Data governance frameworks
- HIPAA-constrained delivery
- Test-driven development for pipelines
- Mentorship & technical guidance
- Multi-engagement delivery
- Level-of-effort estimation
- Discovery workshops
- Stakeholder management
- Agile / Scrum
The person
Outside the work
A hire at this level is someone a team lives with for years, not a résumé. This is who that is.
- Boxing
- Salsa, Bachata, Merengue
- Woodworking
Plus 15 countries across 4 regions, most of them where the music I dance to comes from.
Let’s talk
Send the team and the problem it owns. You will get a straight answer on whether I am the right fit, and the two projects closest to it.