Independent data engineering consulting focused on one thing: building the governed, high-quality data foundations that make AI, analytics, and decisions actually work.
Focused data engineering expertise for organizations that need their data ready for AI, not just for reporting.
Domain-driven data products, lineage, and PII governance that make your data trustworthy enough for AI agents, RAG, and analytics to actually rely on — not just descriptive dashboards.
Legacy ETL re-architecture, cloud data lake migrations, and dbt-based warehouse builds on BigQuery, Snowflake, and AWS — turning brittle pipelines into reliable, well-modeled systems.
AI and agent layers built directly on top of governed data — natural-language analytics, automated insight generation, and decision support that's grounded in data you can trust.
Outcomes from two decades of hands-on data architecture and engineering work.
90%+
Warehouse pipeline reliability achieved, up from 33% — with refresh time cut from 6+ hours to roughly 2.5 hours.
130+
Outdated datasets deprecated after lineage analysis and stakeholder engagement, reducing warehouse clutter and risk.
10TB
Of legacy sensor and log data migrated from CSV to Parquet, powering a from-scratch cloud data lake built on 40,000 lines of refactored Python.
200+
Executive and compliance dashboards designed and shipped on governed, well-modeled data pipelines.
Sierra Data Labs is the independent data engineering consultancy of Culley Harrelson, a data architect and engineer based in the San Francisco Bay Area with over two decades of experience building the data foundations behind analytics, compliance reporting, and now AI.
Across roles as a senior data engineer, cloud data architect, and principal data architect — spanning insurtech, industrial IoT, and enterprise scale at Google — I've led warehouse modernizations, legacy migrations, and the design of an AI-native data platform pairing governed, domain-driven data products with a conversational analytics layer. That work is the thesis behind Sierra Data Labs: data quality is the primary technical goal, because trustworthy data is what makes AI, analytics, and decisions actually work.
dbt Certified Developer · AWS Cloud Practitioner · Certified Scrum Product Owner
"The main technical goal is to create high-quality data."
This principle drives everything I build. Every pipeline, every schema decision, every AI integration is evaluated against whether it makes the data more accurate, more complete, and more useful — because that's what a model, an agent, or a human decision-maker actually needs.
Have an AI initiative stalled on messy data, or a warehouse that needs a rethink? Let's talk.
